Calculating and Acting on the Right KPIs is Critical for Success

Advanced metrics for Customer Acquisition Cost (CAC) and Lifetime Value (LTV)

What is the purpose of CAC, LTV and Payback Period?

Often entrepreneurs we meet have a list of their KPI’s, but as we dig deeper, it becomes clear that they don’t fully comprehend the purpose of the KPI or why we think that they are so critical in helping us understand the health of a business. I’m a believer in thoroughly understanding the economics of your business, and the metrics around Customer Acquisition Cost (CAC), Customer Lifetime Revenue (LTR), Customer Lifetime Value (LTV) and Payback Period are important metrics to work on improving. In this post we use the word customer to mean a ‘buyer’, someone who actually orders product and will use customer and buyer interchangeably.

  • LTR (Customer Lifetime Revenue) is the total revenue from a customer over their estimated lifetime (we prefer using 5 years as representing lifetime)
  • MCA (Marketing Spend for Customer Acquisition) is the marketing cost directed towards new customer acquisition
  • CAC (Customer Acquisition Cost) captures the cost of acquiring a new buyer
  • LTV (Customer Lifetime Value) is the estimated lifetime profits on new customers once they have been acquired
  • Net LTV is the estimated lifetime profits on a new customer taking acquisition cost into account or: LTV – CAC
  • Payback Period is the time for a new customer to generate profits that equals his or her CAC

When LTV is compared to CAC a measure of the future health of the business can be seen. Most VCs will be loath to invest in any company where the ratio of LTV/CAC is less than 3X and enthusiastic when it exceeds 5X. I also favor companies with a short payback period especially if the recovery is from the first transaction as this means almost no cash is consumed when acquiring a customer.

I have previously written about Contribution Margin (gross margin less marketing and sales spend) being one of the most important metrics for companies as it tells me how scaling revenue will help cover G&A and R&D costs. Once Contribution Margin exceeds these costs there is operating profit. A company with low Contribution Margin needs much greater scale to be able to reach profitability than one with high Contribution Margin. Both the LTV/CAC ratio and Payback Period are important predictors of future Contribution Margin levels. Table 1 shows two companies with the same revenue, R&D and G&A cost. Assuming R&D and G&A are fixed for both, at 5% Contribution Margin Company A would need to reach $50 million in revenue to break-even while Company B would be at break-even at under $4 million in revenue given its 65% Contribution Margin.

Break Even Revenue = (R&D + G&A)/ (Contribution Margin %)

For Company A = $2,500,000/ (5%) = $50,000,000

For Company B = $2,500,000/ (65%) = $3,846,154

The problem for Company A stems from the fact that for every $40 spent in acquiring a customer the total LTV (or subsequent profits on the customer) is only $76 making Net LTV $36. With such a low ratio of LTV/CAC, Contribution Margin is also quite low. I’ve chosen two extremes in gross margin to better illustrate the impact of gross margin on Contribution Margin. 

Table 1: Illustrative Example on the Importance of Contribution Margin

In a second example, Table 2, we show the impact of much more efficient marketing and remarketing on a company’s results. We assume gross margin, R&D and G&A are the same for Company C and Company D, but that Company D is much more efficient at acquiring and retaining customers, resulting in a lower CAC and a higher LTR and LTV for Company D. This could be due to higher spending on branding and remarketing, but also because they are very focused on the metrics that help them optimize acquisition and LTR. The result is considerably higher LTV/CAC and significantly higher Contribution Margin. 

Table 2: Impact of Efficient Marketing & Remarketing on a Company’s Results

A few rules regarding Contribution Margin

  1. Higher Gross Margin means greater opportunity for high Contribution Margin – for example, Company A at 20% Gross Margin is severely limited in reaching a reasonable Contribution Margin while Company B at 80% Gross Margin has an opportunity for high Contribution Margin and can reach profitability at lower revenue levels
  2. Companies with many new customers acquired through “free” methods like SEO, viral marketing, etc. have higher Contribution Margins than similar companies with a low proportion of free acquisitions.  Notice Company D can spend less on marketing than Company C because it acquires half of its customers from free sources whereas Company C has none from free sources.
  3. Returning buyers contribute heavily to Contribution Margin since they require little if any marketing cost. A high LTV/CAC ratio usually means customers return more often, which in turn should increase future Contribution Margin. Company D with its very high LTV/CAC likely has much lower churn than Company C and therefore most of its marketing spend is for adding incremental customers rather than just replacing those that churn. If a company simply stopped spending on marketing, its revenue would be from existing customers and new customers acquired from free methods making Contribution Margin roughly equal to Gross Margin. Of course, lack of marketing could have a major impact on future growth
  4. All other things being equal, companies with short Payback Periods tend to have higher Contribution Margin than those with longer Payback Periods

Marketing Spend on Branding versus Customer Acquisition

Branding is the communication of characteristics, values and attributes of an organization and its products. What is the purpose of spending on branding?  To create a strong image of the company and its products so that existing customers want to stay and spending on customer acquisition will be more efficient. Great brands should have a lower CAC and higher LTV than weak brands.

Nike’s “Just Do It” campaign worked well in establishing a brand message that resonated. Their more recent campaign centered around Colin Kaepernick was riskier as it had pictures of the quarterback and the slogan: “Believe in something even if it means sacrificing everything”. While Kaepernick is controversial, he appeals to a large part of the Nike existing and potential buying audience. The campaign helped lift sales by 27% in the first 4 days following the ad launch. Karen McFarlane, founder Kaye Media Partners summed up Nike’s strategy: “Nike’s mission is to bring inspiration and innovation to every athlete in the world. Colin Kaepernick, through his advocacy, conviction, and talent on and off the field, exemplifies those values in the strongest of terms. Couple that with Nike’s commitment to diversity and community, particularly against the backdrop of today’s America where politics have amplified cultural divisions, Nike took the opportunity to lean into their mission and values.”

Marketing campaigns that are for direct customer acquisition differ from branding campaigns as their goal is to directly acquire customers. It is not always clear whether marketing dollars should be classified as acquisition or branding. The key point is that branding is building for the future, whereas acquisition campaigns are measured by how efficiently they deliver current customers. This brings us back to measuring CAC, as the question becomes whether branding campaigns are part of the calculation of CAC. I believe they are not, as CAC should represent the direct cost of acquiring a customer so that increasing that spend will allow us to estimate how many more customers will result. So, I conclude that marketing for branding should be excluded from the calculation of CAC but can be an important strategy to improve efficiency in acquiring customers and maximizing their value. It should be part of the Contribution Margin calculation.

Paid CAC vs Blended CAC

If CAC represents what it costs to acquire a customer, what about those customers that are acquired through SEO, viral marketing, or some other free method? If one simply divides the spend on Marketing for Customer Acquisition (or MCA) by the total number of new customers, the result will be deceptive if it is used to estimate how many more customers will be acquired if MCA spend is increased. Instead companies should calculate CAC in two ways:

Blended CAC = MCA/ (total number of new customers)

Acquisition CAC = MCA/ (number of new customers generated from acquisition marketing only)

Attribution Models and how they help understand CAC by Channel

CAC helps predict the impact of different levels of spending on marketing. Companies also need to know where spending has the best impact, and therefore need to calculate a CAC (and LTV) for each channel of marketing. If a company uses multiple channels like Facebook, Google, and Snap they need to assign credit to each channel for the customers generated from that channel as compared to the spend for that channel to determine the channel CAC. The most simplistic way of doing this is to say, for example, that Google advertising gets credit for the customer if the purchase by the customer occurs after clicking to the site from Google. But what if a consumer sees a Facebook ad, goes to the site to look at products, then a week later comes back to the site from a Snap ad, and then visits the site again because the company sent them an email that they click on, and finally buys something after clicking on a Google ad?  Should Google get the entire credit for the customer?

Attribution models from companies like Amplitude and Hive attempt to appropriately credit each channel touchpoint for the role it played in acquiring a customer. For example, a company might assign 50% of the credit to the last touchpoint and divide the other 50% equally among prior touchpoints or use some other formula that is believed to capture how each channel participated in driving that consumer towards a purchase. Attribution models attempt to help companies understand how scaling spend in each channel will impact customer acquisition. For example:

Google CAC = (Google marketing spend)/ (number of customers attributed to Google ads)

One question that arises when using attribution models to credit each channel is whether “free” areas should share the credit when both paid and free touchpoints occur before converting a consumer into a buyer. The issue is whether the free touchpoints would have occurred had the company not spent on paid marketing. In my prior example the customer would never have received an email from the company had they not first seen the Facebook and Snap ads so eliminating these ads would also eliminate the email which in turn probably means they never would have become a customer. In a similar way, someone who clicked on a Google ad but did not buy, might later do a search for the product, and go to the site from SEO and then buy. So, one method of assigning attribution would be to divide the entire attribution among paid channels if a consumer touched both paid and free ones before becoming a customer. If the consumer only went to free channels, then they would count in blended CAC but not as a customer in calculating paid CAC or CAC by paid channel.

Where does Marketing to an existing customer fit in?

CAC is meant to measure new customer acquisition. So, marketing to an existing customer to get them to buy more should not be part of CAC but rather should be subtracted from MCA. I view it as a cost that reduces Lifetime Value (LTV). Many startups ignore the possibility of “reactivating” churned customers through marketing spend. Several Azure portfolio companies have seen success in deploying a reactivation strategy. I consider this spend as marketing to an existing customer and resulting profits an increase in LTV. Revising the definition of LTV (see Table 1):

LTV = LTR – COGS – Marketing Costs to existing customers

MCA (Marketing Cost for Customer Acquisition) = Marketing Costs – branding costs – marketing to existing customers

In Table 1, remarketing expense is 1% of Lifetime revenue for both Company A and Company B or:

Company A LTV = (GM% – remarketing%) X LTR = (20% – 1%) X $400 = $76

Company B LTV = (80%– 1%) X $400 = $316

What about new customers that return the product?

This is a thorny issue. If a consumer buys their first product from a company and then returns it, is she a customer? I think it would not be wrong with interpreting this either way. But I prefer considering her a customer as she was a buyer, and while the return zeros out the revenue from that purchase, she did become a customer. There is ample opportunity for marketing to her again. Considering her a customer lowers CAC as there are more customers for the spend, but it also lowers LTV for the same reason. So, if the interpretation is applied consistently over both CAC and LTV, I believe it would be correct.

More mature companies should apply the methods shown here to better understand their business

By better understanding acquisition CAC and LTV of each channel of customer acquisition a company can direct more spending to the most efficient ones (those with the highest LTV/CAC). By experimenting with “reactivation” spending a company can determine if this improves LTV/CAC. Companies that improve LTV/CAC will likely generate higher Contribution Margin. Those that increase the proportion of customers acquired for “free” can also improve Contribution Margin. Improving CAC and LTV can be accomplished in several ways as described in prior posts:

Optimizing the cost of customer acquisition modeling metrics to drive startup success (March 2014)

How to improve Contribution Margin (Nov 2018)

Without proper metrics a company is essentially flying blind and will be far less likely to succeed. This post provides a blueprint towards a more sophisticated approach to understanding any business.

Defining Key Elements of the New Model for Retail

In our October, 2015 Soundbytes (https://soundbytes2.com/2015/10/)  I predicted that Omnichannel selling would become prevalent over the ensuing years with brick and mortar retailers being forced to offer an online solution, ecommerce companies needing to access buyers at physical locations and online brands (referred to as DTC or direct to consumer) being carried by 3rd party physical stores. Since that post, these trends have accelerated (including Amazon’s announcement last week that it is opening another “4-star store” in the bay area). Having had more time to observe this progression, I have developed several theories regarding this evolving new world that I would like to share in this post.  

Issues for Brick and Mortar Stores when they Create an Online Presence

Physical retailers are not set up to handle volumes of online sales. Their distribution centers are geared towards sending larger volumes of products to their stores rather than having the technology and know-how to deal directly with consumers (a situation which motivated Walmart to buy Jet for $3.3 billion). In general, a retailer starting to sell online will need to create one or more new distribution centers that are geared towards satisfying direct to consumer online demand.  

Another rude awakening for brick and mortar retailers when they go online is a dramatic increase in returns. On average, returns are about 9% of purchases from a retail store and 30% when purchased online. The discrepancy is even greater for clothing (especially shoes) as fit becomes a major issue. Given that consumers expect free shipping, and most want free return shipping, this becomes a cost that can eviscerate margins. The volume of returns also creates the problem of handling reverse logistics, that is tracking the return, crediting the customer, putting it back into the inventory system as available and restocking it into the appropriate bin location. Then there is the question as to whether the item can still be resold. For clothing this may require adding the cost of cleaning and pressing operations to keep the item fresh and having the systems to track movement of the inventory through this process.

Lastly, the question becomes whether a brick and mortar retailers’ online sales will (at least partly) cannibalize their in-store sales. If so, this, coupled with the growth of online buying, can make existing store footprints too large, reducing store profits.

If Brick and Mortar Retailers Struggle with an Omnichannel Approach, why do DTC Brands Want to Create an Offline Presence?

The answer is a pretty simple one: market access and customer acquisition.  Despite a steady gain of share for online sales, brick and mortar still accounts for over 70% of consumer purchases. Not too long ago, Facebook was a pretty efficient channel to acquire customers. For the past 5 years, Azure portfolio companies have experienced a steep rise in CAC (customer acquisition cost) when using Facebook as the acquisition vehicle. There are many theories as to why, but it seems obvious to me that it is simply the law of supply versus demand. Facebook usage growth has slowed but the demand for ad inventory has increased dramatically, driving up prices. For large brands that use advertising for brand building rather than customer acquisition this does not appear to be a problem, especially when comparing its value to ads on television. For brands that use it for customer acquisition, doubling CAC changes the ratio of LTV (lifetime value or lifetime profits on a customer) to CAC making this method of customer acquisition far less effective.

The combination of these factors has led larger (and smaller) online brands to open brick and mortar outlets. Players like Warby Parker, Casper, Bonobos and even Tesla have done it by creating stores that are a different experience than traditional retail. Warby Parker, Bonobos and Tesla do not stock inventory but rather use the presence to attract customers and enable them to try on/test drive their products. I have bought products, essentially online, while at Warby Parker and Tesla physical locations.

I then had to wait between 2 to 6 weeks for the product to be manufactured and delivered (see the soundbite on Tesla below). What this means in each of their cases is that they kept their business models as ones of “manufacture to demand” rather than build to inventory.  It seems clear that for all four of the companies cited above there is a belief that these physical outlets are a cost effective way of attracting customers with a CAC that is competitive to online ads. They also effectively use online follow-up once you have visited their brick and mortar outlet, thus creating a blend of the two methods. Once the customer is acquired, repeat purchases may occur directly online or in a combination of online and offline.

The Future Blend of Online/Offline

While we have seen a steady progression of companies experimenting with Omnichannel whether they started as offline or online players, we have yet to see an optimal solution. Rather, various players have demonstrated parts of that optimization. So, I’d like to outline a few thoughts regarding what steps might lead to more optimization:

  1. To the degree possible, online purchases by consumers should have an in-store pickup and review option at some savings versus shipping to the home. For clothing, there should also be an opportunity to try the online purchased items on before leaving the store. In that way consumers have the ability to buy online, coupled with the convenience of trying products on in a store. This would expose the customer to a broader set of inventory (online) than even a large footprint store might be able to carry. It would improve fit, lower cost to the brand (by lowering returns and reducing shipping cost) while allowing the brand to begin acquiring better information on fit – insuring an improvement for the next online purchase. A secondary benefit would be the increase in store traffic that was created.
  2. Many retailers will add Rental to the mix of options offered to customers to improve profits. Azure portfolio company, Le Tote, is a subscription rental company for everyday women’s clothes. As women give feedback on a large variety of aspects of fit and preferences it can improve the fit dramatically with each successive box. Retailers need to have systems that replicates this knowledge of their customers. The problem for pure brick and mortar retailers is that they have not had a relationship that enables them to get the feedback…and they don’t have software systems to build this knowledge even if they were to get it. Le Tote has also built up strong knowledge of women’s preferences as to style and has created successful house brands that leverage that knowledge based on massive feedback from subscribers. You may have seen the announcement that Le Tote has just acquired Lord & Taylor, the oldest department store in the country. It plans to use the millions of existing Lord & Taylor customers as a source of potential subscribers to its service. It also has a rental vehicle that can be used to improve monetization of items that don’t sell through at the stores.
  3. Successful online brands will be carried by offline retailers. This has already started to occur but will accelerate over time as DTC brands like Le Tote (and perhaps Stitchfix) use their tens of millions of specific customer feedback data points to produce products that meet the needs expressed in the feedback. If they have correctly mined the data, these brands should be quite successful in offline stores, whether it be their own or a third party retailers’ outlet. The benefit to the retailor in carrying online brands is two-fold: first the online brands that have effectively analyzed their data create products they know can sell in each geography; and second carrying online brands will improve the image of the retailor in the eyes of shoppers who view DTC companies as more forward thinking.
  4. Department store footprints will need to shrink or be shared with online players. The issue discussed earlier of overall ecommerce coupled with brick and mortar stores cannibalizing store demand when they start selling online can be dramatically mitigated by having smaller stores. In that way retailers can maintain their brand presence, continue to get foot traffic, and improve store efficiency.  Any larger footprint store may need to take part of its space and either sublet it (as Macys is doing in some locations) or attract online brands that are willing to pay for a presence in those stores in the form of a percentage of revenue generated or rent. The offer to DTC brands may be to have a pop-up for a set period, or to agree to a longer-term relationship. By working with DTC brands in this way retailers can improve gross margin per square foot (a critical KPI for brick and mortar players) for poorly utilized portions of their store footprints. The secondary benefit to the retailer would be that the online brands will generate additional traffic to the stores. There are already a few startups that are creating a store within a store concept that carries DTC brands. They hope to be the middleman between DTC brands and large retailers/shopping malls making it easier for the DTC brands to penetrate more locations, and easier for the retailer to deal with one new player that will install multiple DTC brands in their locations.
  5. There will be more combinations of online and offline companies merging. By doing that the expertise needed for each area of the business can be optimized. The online companies presumably have better software, logistics and more efficient methods of acquiring online customers. The brick and mortar retailers have greater knowledge of running a physical store, an existing footprint to carry the online brands, locations that allow for delivery to their stores, and a customer base to market to online (reducing the CAC and increasing LTV).
  6. For Omnichannel companies, revenue attribution is complex but becomes essential to managing where dollars are spent. Revenue attribution is the tracking, connecting, and crediting marketing efforts to their downstream revenue creation. For example, if a potential customer responds to a Facebook ad by going online to look at items, then visits a store to check them out live, but eventually buys one or more of the items in response to a google ad, the question becomes: which channel should get credit for acquiring the customer? This is important as the answer may impact company strategy and help determine where marketing dollars get spent. Several Azure portfolio companies are now using 3rd party software from companies like Hive to appropriately give attribution to each channel that helped contribute to the eventual sale. This process is important as it helps determine future spending. We expect better run Omnichannel companies to evolve their analysis of marketing to include attribution models.  

Conclusion: The future winners in retail will be those that successfully migrate to the most optimal omnichannel models

What I have described in this post is inevitable. Some large proportion of customers will always want to do some or all of their shopping at a brick & mortar store. By blending the positive attributes of physical retail with the accessibility to the larger number of options that can exist online, companies can move to more optimal models that address all potential customers. But unless this is done in an intelligent way booby traps like inefficient floor space, excessive returns, high shipping costs and more will rear their ugly heads. This post describes steps for retailers/brands to take that are a starting point for optimizing an omnichannel approach.  

Soundbytes

  • When I had just left Wall Street, I received calls from the press and a very large investor in Hewlett Packard regarding my opinion of the proposed acquisition of Compaq Computer. I said: “HP is in 6 business areas with Imaging being their best and PCs their worst. Doubling up on the worst of the 6 does not make sense to me.” When asked what they should do instead, I replied: “Double up on the best business: acquire Xerox.” My how the tide has turned as Xerox was in trouble then and could have been bought at a very low price. Now it appears Xerox may acquire HP. To be clear, Xerox is still a much smaller market cap company…but I’m enjoying seeing how this process will work out.

  • In the last Soundbytes, I mentioned that I had purchased a Tesla Model 3. What is interesting is that 6 weeks later I am being told that it may take as long as 3-4 more weeks before I receive the car.  This means delivery times have extended to at least 9 weeks. I can’t say how reliable this is but the salesperson I am dealing with told me that Tesla has prioritized production of Model S, Model X and shipments to Europe and Asia over even the more expensive versions of Model 3’s (mine cost almost $59,000 before sales tax). One can easily conclude that production must be at full capacity and that the mix this quarter will contain more higher priced cars. So, demand in the quarter appears to be in excess of 100,000 units and price per car appears strong. Assuming the combination of maximum production in the U.S. and some production out of the Chinese factory, supply might also exceed 100,000 units. If the supply is available, then Tesla should have a strong Q4. However, there is the risk that Tesla doesn’t have the parts to supply both factories or that they have somehow become less efficient. The latest thing to drive down the Tesla stock price is the missteps in showcasing the new truck. I’m not sure why a company should be castigated for an esoteric feature not working in a prototype of a product that won’t be in production until sometime in 2021. Remember, Tesla at its core is a technology company producing next gen autos. I’ve seen other technology companies like Microsoft and Oracle have glitches in demos of future products without such a reaction. As for the design of the truck, I believe Tesla is targeting a 10% to 20% share of the truck market with a differentiated product rather than attempting to attract all potential buyers. 10% of the U.S. pickup market would result in 250,000 units per year. The company has announced pre-orders for the vehicle have already reached 200,000. If these orders are real, they have a home run on their hands but since the deposit is only $100 there is no guarantee that all deposits will convert to actual purchases when the truck goes into production.  

Comparing Recent IPO Companies – Should Performance Drive Valuation?

In a past life, while on Wall Street, one of my favorite calls was: “Buy Dell Short Kellogg”. My reasoning behind the call was that while Dell’s revenue and earnings growth was more than 10X that of Kellogg, somehow Kellogg had a much higher PE than Dell. Portfolio managers gave me various reasons they claimed were logical to explain the un-logical situation like: “Kellogg is more reliable at meeting earnings expectations” …. when in truth they had missed estimates 20 straight quarters. What I later came to believe was that the explanation was their overall comfort level with Kellogg because they understood cereal better than they understood a direct marketing PC hardware company at the time. My call worked out well as Dell not only had a revenue CAGR of nearly 50% from January 1995 (FY 95) through January of 2000 (and a 69% EPS CAGR) but also experienced significant multiple expansion while Kellogg revenue grew at just over 1% annually during the ensuing period and its earnings shrunk (as spend was against missed revenue expectations). The success of Dell was a major reason I was subsequently selected as the number one stock picker across Wall Street analysts for 2 years in a row.

I bring up history because history repeats. One of the reasons for my success in investing is that I look at metrics as a basis of long-term valuation. This means ignoring story lines of why the future is much brighter for those with weak metrics or rationales of why disaster will befall a company that has strong results. Of course, I also consider the strength of management, competitive advantages and market size. But one key thesis that comes after studying hundreds of “growth” companies over time is that momentum tends to persist, and strong business models will show solid contribution margin as an indicator of future profitability.

Given this preamble I’ll be comparing two companies that have recently IPO’d. Much like those that supported Kellogg, the supporters of the one with the weaker metrics will have many reasons why it trades at a much higher multiple (of revenue and gross margin dollars) than the one with stronger metrics.  

Based on financial theory, companies should be valued based on future cash flows. When a company is at a relatively mature stage, earnings and earnings growth will tend to be the proxy used and a company with higher growth usually trades at a higher multiple of earnings.   Since many companies that IPO have little or no earnings, many investors use a multiple of revenue to value them but I prefer to use gross margin or contribution margin (where marketing cost is broken out clearly) as a proxy for potential earnings as they are much better indicators of what portion of revenue can potentially translate to future earnings (see our previous post for valuation methodology).

I would like to hold off on naming the companies so readers can look at the metrics with an unbiased view (which is what I try to do). So, let’s refer to them as Company A and Company B. Table 1 shows their recent metrics.

Notes:

  • Growth for Company B included an extra week in the quarter. I estimate growth would have been about 27% year/year without the extra week
  • Disclosures on marketing seem inexact so these are estimates I believe to be materially correct
  • Pre-tax income for Company A is from prior quarter as the June quarter had considerable one-time expenses that would make it appear much worse

Company B is:

  1. Growing 2 -2 ½ times faster
  2. Has over 3X the gross margin percentage
  3. Over 28% contribution margin whereas Company A contribution margin is roughly at 0
  4. Company B is bordering on profitability already whereas Company A appears years away

Yet, Company A is trading at roughly 3.5X the multiple of revenue and almost 11X the multiple of gross margin dollars (I could not use multiple of contribution margin as Company A was too close to zero). In fairness to Company A, its gross margin was much higher in the prior quarter (at 27%). But even giving it the benefit of this higher number, Company B gross margins were still about 65% higher than Company A.

The apparent illogic in this comparison is much like what we saw when comparing Kellogg to Dell many years ago. The reasons for it are similar: investors, in general, feel more comfortable with Company A than they do with Company B. Additionally, Company A has a “story” on why things will change radically in the future. You may have guessed already that Company A is Uber. Company B is Stitchfix, and despite its moving to an industry leading position for buying clothing at home (using data science to customize each offering) there continues to be fear that Amazon will overwhelm it sometime in the future. While Uber stock has declined about 35% since it peaked in late June it still appears out of sync compared to Stitchfix.

I am a believer that, in general, performance should drive valuation, and have profited greatly by investing in companies that are growing at a healthy rate, appear to have a likelihood of continuing to do so in the near future and have metrics that indicate they are undervalued.

Soundbytes

  • It appears that many others are now beginning to focus more closely on gross margins which we have been doing for years. I would encourage a shift to contribution margin, where possible, as this considers the variable cost needed to acquire customers.
  • A few notes about Tesla following our 2019 predictions: My household is about to become a two Tesla family. My wife has owned her second Tesla, a Model S, for over 4 years and I just placed an order to buy a Model 3 as a replacement for my Mercedes 550S. Besides the obvious benefits to the environment, I’m also tired of having to go to gas stations every week. The Model 3 can go 310 miles on a charge, is extremely fast, has a great user interface and has autopilot. I looked at several other cars but found it hard to justify paying twice as much (or more) for a car with less pickup, inferior electronics, etc.
  • If you were wondering why Tesla stock has gone on a run it is because the Chinese Ministry of Industry and Information Technology (MIIT) has added Tesla to its list of approved auto manufacturers (the news of the possibility broke over a week ago). It appears likely that Tesla will begin producing Model 3s out of the new Giga Factory in China some time in Q4. This not only adds capacity for Tesla to increase its unit sales substantially in 2020 but also will save the Company considerable money as it won’t need to ship cars from its US factory. Remember Tesla also is planning on a Giga Factory in Europe to service strong demand there. The company has said that it will choose the location by the end of 2019. Given the intense competition to be the selected location, it is likely that the site chosen will involve substantial incentives to Tesla. While I would not want to predict when it will be in production, Elon Musk expects the date to be sometime in 2021. Various announcements along the way could be positive for Tesla stock.

Why Apple Acquiring Tesla Seems an Obvious Step…

…and why the obvious probably won’t happen!

A Look at Apple history

Apple’s progress from a company in trouble to becoming the first company to reach a trillion dollar market cap meant over 400X appreciation in Apple stock. The metamorphosis began when the company hired Fred Anderson as an Executive VP and CFO in 1996. Tim Cook joined the company as senior VP of worldwide operations in 1998. Fred and Tim improved the company operationally, eliminating wasteful spending that preceded their tenure. Of course, as most of you undoubtedly know, bringing back Steve Jobs by acquiring his company, NeXT Computer in early 1997 added a strategic genius and great marketer to an Apple that now had an improved business model. Virtually every successful current Apple product was conceived while Steve was there. After Fred retired in 2004, Tim Cook assumed even more of a leadership role than before and eventually became CEO shortly before Jobs’ death in 2011.  

Apple post Steve Jobs

Tim Cook is a great operator. In the years following the death of Steve Jobs he squeezed every bit of profit that is possible out of the iPad, iPod, iMacs, music content, app store sales and most of all the iPhone. Because great products have a long life cycle they can increase in sales for many years before flattening out and then declining.

Table 1: Illustrative Sales Lifecycle for Great Tech Product

Cook’s limit is that he cannot conceptualize new products in the way Steve Jobs did. After all, who, besides an Elon Musk, could? The problem for Apple is that if it is to return to double digit growth, it needs a really large, successful new product as the iPhone is flattening in sales and the Apple Watch and other new initiatives have not sufficiently moved the needle to offset it. Assuming Q4 revenue growth in FY 2019 is consistent with the first 9 months, then Apple’s compound growth over the 4 years from FY 15 to FY 19 will be 3.0% (see Table 2) including the benefit of acquisitions like Beats.

iPhone sales have flattened

The problem for Apple is that the iPhone is now in the mature part of its sales life cycle. In fact, unit sales appear to be declining (Graph 1) but Apple’s near monopoly pricing power has allowed it to defy the typical price cycle for technology products where average selling prices decline over time. The iPhone has gone from a price range of $99 to $299 in June 2009 to $999 to $1449 for the iPhoneX, while the older iPhone 7 is still available with minimal storage for $449. That’s a 4.5X price increase at the bottom and nearly 5X at the high end! This defies gravity for technology products.

Graph 1: iPhone Unit Sales (2007-2018)

In the many years I followed the PC market, it kept growing until reaching the following set of conditions (which the iPhone now also faces):

  1. Improvements in features were no longer enough to drive rapid replacement cycles
  2. Pricing was under pressure as component costs declined and it became more difficult to convince buyers to add capacity or capability sufficient to hold prices where they were
  3. The number of first time users available to buy product was no longer increasing each year
  4. Competition from lower priced suppliers created pricing pressure

Prior to that time PC pricing could be maintained by convincing buyers that they needed one or more of:

  1. The next generation of processor
  2. A larger or thinner screen
  3. Next generation storage technology

What is interesting when we contrast this with iPhones is that PC manufacturers struggled to maintain average selling prices (ASPs) until they finally began declining in the early 2000s. Similarly, products like DVD players, VCRs, LCD TVs and almost every other technology driven product had to drop dramatically in price to attract a mass market. In contrast to that, Apple has been able to increase average prices at  the same time that the iPhone became a mass market product. This helped Apple postpone the inevitable revenue flattening and subsequent decline due to lengthening replacement cycles and fewer first time buyers. In the past few years, other then the bump in FY 2018 from the launch of the high priced Model X early that fiscal year, iPhone revenue has essentially been flat to down. Since it is well over 50% of Apple revenue, this puts great pressure on overall revenue growth.

To get back to double digit growth Apple needs to enter a really large market

To be clear, Apple is likely to continue to be a successful, highly profitable company for many years even if it does not make any dramatic acquisitions. While its growth may be slow, its after tax profits has been above 20% for each of the past 5 years. Strong cash flow has enabled the company to buy back stock and to support increasing dividends every year since August 2014.

Despite this, I think Apple would be well served by using a portion of their cash to make an acquisition that enables them to enter a very large market with a product that already has a great brand, traction, and superior technology. This could protect them if the iPhone enters the downside of its revenue generating cycle (and it is starting to feel that will happen sometime in the next few years). Further, Apple would benefit if the company they acquired had a visionary leader who could be the new “Steve Jobs” for Apple.

There is no better opportunity than autos

If Apple laid out criteria for what sector to target, they might want to:

  1. Find a sector that is at least hundreds of billions of dollars in size
  2. Find a sector in the midst of major transition
  3. Find a sector where market share is widely spread
  4. Find a sector ripe for disruption where the vast majority of participants are “old school”

The Automobile industry matches every criterion:

Matching 1.  It is well over $3 trillion in size

Matching 2. Cars are transitioning to electric from gas and are becoming the next technology platform

Matching 3. Eight players have between 5% and 11% market share and 7 more between 2% and 5%

Matching 4. The top ten manufacturers all started well over 50 years ago

And no better fit for Apple than Tesla

Tesla reminds me of Apple in the late 1990s. Its advocates are passionate about the company and its products. It can charge a premium versus others because it has the best battery technology coupled with the smartest software technology. The company also designs its cars from the ground up, rather than retrofitting older models, focusing on what the modern buyer would most want. Like Jobs was at Apple, Musk cares about every detail of the product and insists on ease of use wherever possible. The business model includes owning distribution outlets much like Apple Stores have done for Apple. By owning the outlets, Tesla can control its brand image much better than any other auto manufacturer. While there has been much chatter about Google and Uber in terms of self-driving cars, Tesla is the furthest along at putting product into the market to test this technology.

Tesla may have many advantages over others, but it takes time to build up market share and the company is still around 0.5% of the market (in units). It takes several years to bring a new model to market and Tesla has yet to enter several categories. It also takes time and considerable capital to build out efficient manufacturing capability and Tesla has struggled to keep up with demand. But, the two directions that the market is moving towards are all electric cars and smart, autonomous vehicles. Tesla appears to have a multi-year lead in both. What this means is that with enough capital and strong operational direction Tesla seems poised to gain significant market share.

Apple could accelerate Tesla’s growth

If Apple acquired Tesla it could:

  1. Supply capital to accelerate launch of new models
  2. Supply capital for more factories
  3. Increase distribution by offering Tesla products in Apple Stores (this would be done virtually using large computer screens). An extra benefit from this would be adding buzz to Apple stores
  4. Supply operational knowhow that would increase Tesla efficiency
  5. Add to the luster of the Tesla brand by it being part of Apple
  6. Integrate improved entertainment product (and add subscriptions) into Tesla cars

These steps would likely drive continued high growth for Tesla. If, with this type of support, it could get to 5% share in 3-5 years that would put it around $200 billion in revenue which would be higher than the iPhone is currently. Additionally, Elon Musk is possibly the greatest innovator since Steve Jobs. As a result, Tesla would bring to Apple the best battery technology, the strongest power storage technology, and the leading solar energy company. More importantly, Apple would also gain a great innovator.

The Cost of such an acquisition is well within Apple’s means

At the end of fiscal Q3, Apple had about $95 billion in cash and equivalents plus another $116 billion in marketable securities. It also has averaged over $50 billion in after tax profits annually for the past 5 fiscal years (including the current one). Tesla market cap is about $40 billion. I’m guessing Apple could potentially acquire it for less than $60 billion (which would be a large premium over where it is trading). This would be easy for Apple to afford and would create zero dilution for Apple stockholders.

If the Fit is so strong and the means are there, why won’t it happen?

I can sum up the answer in one word – ego.  I’m not sure Tim Cook is willing to admit that Elon would be a far better strategist for Apple than him. I’m not sure he would be willing to give Elon the role of guiding Apple on the product side. I’m not sure Elon Musk is willing to admit he is not the operator that Tim Cook is (remember Steve Jobs had to find out he needed the right operating/financial partners by getting fired by Apple and essentially failing at NeXT). I’m not sure Elon is willing to give up being the CEO and controlling decision-maker for his companies.

So, this probably will never happen but if it did, I believe it would be the greatest business powerhouse in history!

Soundbytes

  1. USA Today just published a story that agreed with our last Soundbytes analysis of why Klay Thompson is underrated.
  2. I expect Zoom Video to beat revenue estimates of $129 million to $130 million for the July Quarter by about $5 million or more

The Warriors Ain’t Dead Yet!

Why the team is still a contender

My long term readers know that every so often the blog wanders into the sports arena. In doing so, I apply the same type of analysis that I do for public stocks and for VC investments to sports, and usually, basketball. Given all the turmoil that has occurred in the NBA this off-season, including the Warriors losing Durant, Iguodala, Livingston, Cousins and several other players, I thought it would be interesting to evaluate the newly changed team. Both ESPN and CBS power rankings have them 7th in the West and 11th in the NBA. I find that an overreaction as the Warriors may have beaten the Raptors if Klay Thompson not been injured, they swept Portland, and won the last 2 Rockets games without Durant. At the time this drove a lot of chatter that the team might be better off without Durant (I disagree).

But rather then compare the revised roster to last year’s, it seems more closely matched with the 2014-15 team, as that was a championship team that did not include Kevin Durant. I will make 2 key assumptions:

  1. Klay Thompson will return by the end of February and be as effective as he was before his injury
  2. The Warriors will make the playoffs despite missing Thompson for the majority of the season

It all starts with Curry

A third key assumption that has been proven over and over again is that players that come to the Warriors usually perform better as they benefit from the “Curry Effect”, namely, getting more shots without having someone closely guarding them, (The Curry Effect), resulting in an average improved shooting percentage of over 5%. In all fairness, it really is the “Curry plus Thompson Effect” as the extreme focus on preventing the two of them from taking 3 point shots is what frees up others. It helps that Curry is unselfish and readily passes the ball when he is double or triple teamed. Thompson’s passing has improved substantially but since he gets his shot off so quickly, he has less need to pass it. Last year both shot over 40% from 3 despite defensive efforts focused on preventing each of them from taking those shots.

Starting Teams: 2019-20 vs 2014-15

Table 1

Curry, Thompson and Green, the heart and soul of the Warriors, all remain from the 2014-15 roster, and now are at their peaks. In the 2014-15 season when Green first became a starter, Curry was one year away from reaching his peak and Thompson was just coming into his own especially on defense. I believe each of them is better today then they were at that time. At his best, Bogut may have been better than Cauley-Stein, but by 2014 Bogut had been through a number of injuries. Last year Cauley-Stein averaged nearly double the points of 2014 Bogut (11.9 vs 6.3), took slightly more rebounds per game and was a better free throw shooter. Stein, much like Bogut, is also considered a solid pick setter and defender. Russell is someone who should benefit greatly from playing with Curry. Even without that, last season he averaged over twice as many points per game as 2014 Barnes (21.1 vs 10.1) which should take considerable pressure off Curry (and Klay when he returns). However, Barnes was a better defender in 2014 than Russell is today. I give the edge to all 5 starters on the 2019 starting team compared to the 5 that started in 2014-15.

Thompson may be the most underrated player in the league!

It’s unfortunate that Thompson was injured in game 6 of the 2019 finals as he was once again proving just how good he can be. Not only was he playing lockdown defense, but he also drove the offense in what has been referred to as a typical Klay game 6. In just 32 minutes, before getting injured, he scored 30 points on 83% effective shooting percentage (67% on 3s), went 10 for 10 on free throws, and had 5 rebounds and 2 steals. I believe Golden State, even without Durant, would have forced a game 7 if Thompson did not get injured.

It boggles my mind that one of the websites could refer to Thompson as “an average player” who did not merit a max contract. This is bordering on the ridiculous and has a lot to do with the fact that the most important measure of shooting, effective shooting percentage (where each 3 made counts as 1½ 2 point shots made) does not normally get reported (or even noticed). In Table 2, I list the top 31 scorers from last season (everyone who averaged at least 20 points per game) and rank them by effective shooting. Thompson is number 8 in effective shooting and number 3 in 3-point percentage among this group. So, if effective shooting percentage was regularly published, Thompson would show up consistently helping the perception of his value. When this is coupled with his being a third team all-defensive player (i.e., one of the top 15 defenders in the league) it appears clear that he should be considered one of the top 15 players in the league.

Table 2: Top Scorers 2018-2019 Season

6th Man 2019-20 vs 2014-15

Kevon Looney has emerged as a potential star in the works. While he may not yet be the defensive presence of Iguodala, he is getting close. His scoring per minute played was higher than Andre’s 2014-15 numbers and his rebounds per minute were more than twice as much. While Iguodala had greater presence and could run the team as well as assist others in scoring, Looney at least partly makes up for this in his ability to set screens. Looney also has a much higher effective shooting percentage (62.7% vs 54.0%) than Andre had in 2014-15. While Kevon doesn’t shoot 3s he gets many points by putting back offensive rebounds and dunking lob passes. Overall, I give the edge to Iguodala based on the Looney of last season but given Looney’s potential to improve this might be dead even in the coming one.

Rest of the Bench for the 2 teams

It is the bench that is hardest to evaluate. Unlike last year’s bench, the 2014-15 bench was quite strong which spawned the Warrior logo “Strength in Numbers”. It included quality veteran players like Leandro Barbosa, David Lee, Mareese Sprouts and Shaun Livingston, who was playing at a much higher level than last season. The four of these together averaged over 26 points per game.  This coming year’s bench is much younger and more athletic. It includes Alec Burks, Glenn Robinson, Alfonzo McKinnie and Omari Spellman, plus several rookies and Jacob Evans III. The first four are all capable of scoring and are solid 3-point shooters (they could increase to well above average once with the Warriors). I expect that group, coupled with one or two of the others, to exceed the 2014-15 bench in defense…but may not have as much scoring fire power. The team is likely to give one or two of the rookies as well as Evans opportunities to earn minutes as well. The bench is an improvement over last year but may not be as strong as the 2014-15 squads.

Overall Assessment

I believe the 2019-2020 squad is better than the championship team of 2015. The starting lineup features the core 3 players who have improved since then, D’Angelo Russell who was an all-star last year, and a solid center in Willie Cauley-Stein making the edge substantial. Looney as 6th man is already giving evidence of future stardom. While he was not quite the Andre Iguodala of 2014-15, the difference is modest, and Looney continues to improve. The 2014-15 bench appears superior to that of next season, but the edge is not great as the newer group should be stronger defensively and is not far off the older group as scorers – the question will be how well they gel and how much the Curry/Klay factor improves their scoring. Finally, I think Kerr is a better coach today than he was given the last 5 years of experience.

They May Have Improved vs 2014-15, but so has the Competition

ESPN and CBS power rankings reflect the fact that multiple teams have created new “super star” two-somes:

  • Lakers: Lebron and Anthony Davis
  • Clippers: Kawhi Leonard and Paul George
  • Houston: Harden and Westbrook (in place of Chris Paul)
  • Nets: Kyrie Irving and Durant

Contenders also include improving young teams like Boston, Philadelphia, Denver, and Utah plus an improved Portland squad. This makes the landscape much tougher than when the Warriors won their 2015 championship. Yet, none of these teams seem better than the Cleveland team (led by a younger LeBron, Kyrie Irving and Kevin Love) the Warriors beat in 2015. So, assuming Klay returns by late February and is back to par, I believe the Warriors will remain strong contenders given the starting team with four all-stars augmented by Willie Cauley-Stein and an improving Kevon Looney as 6th man. However, it will be much tougher going in the early rounds in the playoffs, making getting to the finals longer odds than in each of the past 5 years.

SoundBytes

  1. An examination of Table 2 reveals several interesting facts:
  2. Curry, once again is the leader among top scorers in effective shooting and the only one over 60%
  3. Antetokounmpo is only slightly behind despite being a very poor 3-point shooter. If he can improve his distance shooting, he may become unstoppable
  4. Russell Westbrook, once again, had the worst effective shooting percent of anyone who averaged 20 points or more. In fact, he was significantly below the league average. Part of the reason is despite being a very poor 3-point shooter he continues to take too many distance shots. Whereas most players find that taking 3s increases their effective shooting percent, for Westbrook it lowers it. I haven’t been able to check this, but one broadcaster stated that he has the lowest 3-point percentage of any player in history that has taken over 2500 3-point shots!
  5. I believe that Westbrook has a diminished chance to accumulate as many triple doubles next season as he has in the past. There is only one ball and both he and Harden tend to hold it most of the time. When Chris Paul came to the Rockets his assists per game decreased by about 15% compared to his prior 3 season average.

My Crazy Investment Technique for Solid Growth Stocks

You should not try it!

Applying Private Investment Analysis to the Rash of Mega-IPOs Occurring

The first half of 2019 saw a steady stream of technology IPOs. First Lyft, then Uber, then Zoom, all with different business models and revenue structures. As an early investor in technology companies, I spend a lot of time evaluating models for Venture Capital, but as a (recovering) investment analyst, I also like to take a view around how to structure a probability weighted investment once these companies have hit the public markets. The following post outlines a recent approach that I took to manage the volatility and return in these growth stocks.

Question: Which of the Recent technology IPOs Stands out as a Winning Business Model?

Investing in Lyft and Uber, post IPO, had little interest for me. On the positive side, Lyft revenue growth was 95% in Q1, 2019, but it had a negative contribution margin in 2018 and Q1 2019.  Uber’s growth was a much lower 20% in Q1, but it appears to have slightly better contribution margin than Lyft, possibly even as high as 5%. I expect Uber and Lyft to improve their contribution margin, but it is difficult to see either of them delivering a reasonable level of profitability in the near term as scaling revenue does not help profitability until contribution margin improves. Zoom Video, on the other hand, had contribution margin of roughly 25% coupled with over 100% revenue growth. It also seems on the verge of moving to profitability, especially if the company is willing to lower its growth target a bit.

Zoom has a Strong Combination of Winning Attributes

There is certainly risk in Zoom but based on the momentum we’re seeing in its usage (including an increasing number of startups who use Zoom for video pitches to Azure), the company looks to be in the midst of a multi-year escalation of revenue. Users have said that it is the easiest product to work with and I believe the quality of its video is best in class. The reasons for Zoom’s high growth include:

  1. Revenue retention of a cohort is currently 140% – meaning that the same set of customers (including those who churn) spend 40% more a year later. While this growth is probably not sustainable over the long term, its subscription model, based on plans that increase with usage, could keep the retention at over 100% for several years.
  2. It is very efficient in acquiring customers – with a payback period of 7 months, which is highly unusual for a SaaS software company. This is partly because of the viral nature of the product – the host of the Zoom call invites various people to participate (who may not be previous Zoom users). When you participate, you download Zoom software and are now in their network at no cost to Zoom. They then offer you a free service while attempting to upgrade you to paid.
  3. Gross Margins (GMs) are Software GMs – about 82% and increasing, making the long-term model likely to be quite profitable
  4. Currently the product has the reputation of being best in class (see here) for a comparison to Webex.
  5. Zoom’s compression technology is well ahead of any competitor according to my friend Mark Leslie (a superb technologist and former CEO of Veritas).

The Fly in the Ointment: My Valuation Technique shows it to be Over Valued

My valuation technique, published in one of our blog posts, provides a method of valuing companies based on revenue growth and gross margin. It helps parse which sub-scale companies are likely to be good investments before they reach the revenue levels needed to achieve long term profitability. For Zoom Video, the method shows that it is currently ahead of itself on valuation, but if it grows close to 100% (in the January quarter it was up 108%) this year it will catch up to the valuation suggested by my method. What this means is that the revenue multiple of the company is likely to compress over time.

Forward Pricing: Constructing a Way of Winning Big on Appreciation of Even 10%

So instead of just buying the stock, I constructed a complex transaction on May 29. Using it, I only required the stock to appreciate 10% in 20 months for me to earn 140% on my investment. I essentially “pre-bought” the stock for January 2021 (or will have the stock called at a large profit). Here is what I did:

  1. Bought shares of stock at $76.92
  2. Sold the same number of shares of call options at $85 strike price for $19.84/share
  3. Sold the same number of shares of put options at $70 strike for $22.08/share
  4. Both sets of options expire in Jan 2021 (20 months)
  5. Net out of pocket was $35/share

Given the momentum I think there is a high probability (75% or so) that the revenue run rate in January 2021 (when options mature) will be over 2.5x where it was in Q1 2019. If that is the case, it seems unlikely that the stock would be at a lower price per share than the day I made the purchase despite a potential for substantial contraction of Price/Revenue.

In January 2021, when the options expire, I will either own the same shares, or double the number of shares or I will have had my shares “called” at $85/share.

The possibilities are: 

  • If the stock is $85 or more at the call date, the stock would be called, and my profit would be roughly 140% of the net $35 invested
  • If the stock is between $70 and $85, I would net $42 from the options expiring worthless plus or minus the change in value from my purchase price of $76.92. The gain would exceed 100%
  • If the stock is below $70, I’ll own 2x shares at an average price of $52.50/share – which should be a reasonably good price to be at 20 months out.
  •  Of course, the options can be repurchased, and new options sold during the time period resulting in different outcomes.

Break-Even Point for the Transaction Is a 32% Decline in Zoom Video Stock Price

Portfolio Managers that are “Value Oriented” will undoubtedly have a problem with this, but I view this transaction as the equivalent of a value stock purchase (of a high flyer) since the break-even of $52/share should be a great buy in January 2021. Part of my reasoning is the downside protection offered: where my being forced to honor the put option would mean that in January 2021, I would own twice the number of shares at an average price of $52.50/share. If I’m right about the likelihood of 150% revenue growth during the period, it would mean price/revenue had declined about 73% or more. Is there some flaw in my logic or are the premiums on the options so high that the risk reward appears to favor this transaction?

I started writing this before Zoom reported their April quarter earnings, which again showed over 100% revenue growth year/year. As a result, the stock jumped and was about $100/share. I decided to do a similar transaction where my upside is 130% of net dollars invested…but that’s a story for another day.

Estimating the “Probabilistic” Return Using My Performance Estimates

Because I was uncomfortable with the valuation, I created the transaction described above. I believe going almost 2 years out provides protection against volatility and lowers risk. This can apply to other companies that are expected to grow at a high rate. As to my guess at probabilities:

  1. 75% that revenue run rate is 2.5x January 2019 (base) quarter in the quarter ending in January 2021. A 60% compound annual growth (CAG) for 2 years puts the revenue higher (they grew over 100% in the January 2019 quarter to revenue of $105.8M)
  2. 95% that revenue run rate is over 2.0X the base 2 years later (options expire in January of that year). This requires revenue CAG of 42%. Given that the existing customer revenue retention rate averaged 140% last year, this appears highly likely.
  3. 99% that revenue is over 1.5X the base in the January 2021 quarter (requires slightly over 22% CAG)
  4. 1% that revenue is less than 1.5X

Assuming the above is true, I believe that when I did the initial transaction the probabilities for the stock were (they are better today due to a strong April quarter):

  1. 50% that the stock trades over 1.5X today by January 2021 (it is almost there today, but could hit a speed bump)
  2. 80% that the stock is over $85/share (up 10% from when I did the trade) in January 2021
  3. 10% that the stock is between $70 and $85/share in January 2021
  4. 5% that the stock is between $52 and $70 in January 2021
  5. 5% that the stock is below $52

Obviously, probabilities are guesses since they heavily depend on market sentiment, whereas my revenue estimates are more solid as they are based upon analysis, I’m more comfortable with. Putting the guesses on probability together this meant:

  1. 80% probability of 140% profit = 2.4X
  2. 10% probability of 100% profit = 2.0X
  3. 5% probability of 50% profit (this assumes the stock is in the middle at $61/share) = 1.5X
  4. 5% probability of a loss assuming I don’t roll the options and don’t buy them back early. At $35/share, loss would be 100% = (1.0X)

If I’m right on these estimates, then the weighted probability is 120% profit. I’ve been doing something similar with Amazon for almost 2 years and have had great results to date. I also did part of my DocuSign buy this way in early January. Since then, the stock is up 27% and my trade is ahead over 50%. Clearly if DocuSign (or Amazon or Zoom) stock runs I won’t make the same money as a straight stock purchase would yield given that I’m capped out on those DocuSign shares at slightly under 100% profit, but the trade also provides substantial downside protection.

Conclusion: Investing in Newly Minted IPOs of High Growth Companies with Solid Contribution Margins Can be Done in a “Value Oriented” Way  

When deciding whether to invest in a company that IPOs, first consider the business model:

  • Are they growing at a high rate of at least 30%?
  • Experiencing increasing contribution margins already at 20% or more?
  • Is there visibility to profitability without a landscape change?

Next, try to get the stock on the IPO if possible. If you can’t, is there a way of pseudo buying it at a lower price? The transaction I constructed may be to complex for you to try and carries the additional risk that you might wind up owning twice the number of shares. If you decide to do it make sure you are comfortable with the potential future cash outlay.

R&D: Amazon’s Dirty Little Secret Weapon

 

Why doesn’t Amazon produce more earnings given its dominance?

Amazon just reported earnings and, as was the case in 2017 and 2016, emphasized that 2019 will be an investment year, so the strong operating margin expansion of 2018 would be capped in 2019. This, of course, is great fodder for bears on the stock as Amazon gave sceptics renewed opportunity to point out that it is a company that has a flawed business model and would find it difficult to ever earn a reasonable return on revenue.

In contrast, I believe that Amazon continues to transform itself into a potential strong profit performer. For example, taking the longer perspective, Amazon’s gross margins are now over 40% up from 27.2% five years ago (2013). So why doesn’t Amazon deliver higher operating margin than the slightly over 6% it reported in 2018? Amazon’s dirty little secret is that it continues to invest heavily in creating future dominance through R&D. Had it spent a similar amount in R&D to its long time competitor, Walmart, EBITDA would have nearly tripled… to over 17% of revenue! I must confess that in the past I haven’t paid enough attention to how much Amazon spends on R&D. As a result, I was surprised that Apple and Microsoft trailed it in voice recognition technology and that Amazon could lead IBM and Microsoft in cloud technology. The reason this occurred is not a surprising one: Amazon outspends Apple, Microsoft and IBM in R&D.

In fact, Amazon now outspends every company in the world (see Table 1) and have been dedicating a larger portion of available dollars to R&D (as measured by the % of gross margin dollars spent on R&D) than any other large technology company, except Qualcomm, for more than 10 years. Even though Amazon had less than 50% of Apple’s revenue and less than 1/3 of its gross margin dollars 5 years ago (2013) Amazon spent nearly 50% more than Apple on R&D that year… by 2018 the gap had increased to close to 100% more.

Table 1: Top 10 (and a few more) U.S. R&D Spenders in 2018 ($Bn)

Sources: market watch, analyst reports, annual filings

Note 1: Ford and GM may be in the top 10 but so far have not reported R&D in 2018. If they report it at year end the table could change. Walmart does not report R&D and their spend is generally unavailable, but I found a reference that said they expected to spend $1.1M in 2017.

Note 2: A 2018 global list would include auto makers VW and Toyota (with R&D of $15.8B and about 10.0B), drug company Roche (&10.8B) and tech company Samsung at $15.3B in place of the lowest 4 in Table 1.

The Innovators Financial Dilemma: Increasing Future Prospects can lower Current Earnings

When I was on Wall Street covering Microsoft (and others) Bill Gates would often point out that the company was going to make large investments the following year so they could stay ahead of competition. He said he was less concerned with what that meant for earnings. That investment helped drive Microsoft to dominance by the late 1990s. Companies are often confronted with the dilemma of whether to increase spending to drive future growth or to maximize current earnings. I believe that investment in R&D, when effective, is correlated to future success.

It is interesting to see how leaders in R&D spending have transitioned over the past 10 years. In 2008 the global leaders in R&D spending included 5 pharma companies, 3 auto makers and only 2 tech companies (Nokia and Microsoft which subsequently merged). In 2018, 6 of the top 7 spenders (Samsung plus the 5 shown in Table 1) were technology companies.

Table 2 – 2008 global R&D leaders ($Bn)

Note: *Facebook data from 2009, first available financials from S-1 filing

It’s hard to change without tanking one’s stock

When a company has a business model that allocates 1% of gross margin dollars to R&D, it is not easy to turn on the dime. If Walmart had decided to invest half as much as Amazon in R&D in 2018, its earnings would have decreased by 80% – 90% and its stock would have depreciated substantially. So, instead it began a buying binge several years ago to try to close the technology gap through acquisitions (which has a much smaller impact on operating margins). It remains to be seen if this strategy will succeed going forward but in the past 5 years Walmart revenue (including acquisitions) increased only 5% while Amazon’s was up 130% in the same period (also including acquisitions).

Whatever Happened to IBM?

When I was growing up, I thought of IBM as the king of tech. In the early 1990s it still seemed to rule the roost. The biggest fear for Microsoft was that IBM could overwhelm it, yet now it appears to be an also ran in technology. From 2014 to 2018, a heyday era for tech companies, its revenue shrank from $93 billion in 2014 to $80 billion in 2018. I can’t tell how much of the problem stems from under investing in R&D versus poor execution, but for the past 5 years it has spent an average of about 13% of GM on R&D, while the 6 tech companies in Table 1 have averaged about 24% of GM dollars with Apple the only one under 20%.

 

Soundbytes

Soundbyte I: Tesla

  • I recently had a long dialogue with a very smart fund manager and was struck by what I believe to be misinformation he had read regarding Tesla. There were 3 major points that he had heard:
    • The quality of Tesla cars was shoddy
    • Tesla could not maintain reasonable margins as it began producing lower priced Model 3s
    • The upcoming influx of electric cars from companies like Porsche, Jaguar and Audi would take substantial market share away from Tesla

I decided to do a bit of research to determine how valid each of these issues might be.

  • Tesla Quality: I found it hard to believe that the majority of Tesla owners thought the car was of poor quality since every one of the 15 or so people I knew who had bought one had already bought another or were planning to for their next car. So, I found a report on customer satisfaction from Consumer Reports, and I was not surprised to find that Tesla was the number 1 ranked car by customer satisfaction.
  • Tesla margins: this is much harder to predict. Since Tesla is relatively young as a manufacturer it has had numerous issues with production. Yet it is probably ahead of many others when it comes to automating its facilities. This tends to cause gross margins to be lower while volume ramps and higher subsequently. The combination of that, plus moving up the learning curve, should mean that Tesla lowers the cost of producing its products. However, Tesla charges more for cars with higher capacity for distance, but as I understand it uses software to limit battery capacity for lower priced cars. This would mean that a portion of the difference between a lower priced Model 3 and a higher priced one (the battery capacity) would be minimal change in cost, putting pressure on margins. The question becomes whether Tesla’s improving cost efficiencies offset the average price decline of a Model 3 as Tesla begins fulfilling demand for lower priced versions.
  • March 1 Update: After this post was complete (Thursday February 28) the company announced it was closing many showrooms to reduce costs. Then late today (Friday) announced that the $35,000 version of the model 3 is now available. So, we shall soon see the impact. I believe that if Tesla has increased capacity there will be very strong sales. It also likely will experience lower gross margin percentages as it climbs the learning curve and ramps production.
  • Will the influx of electric cars from others impact Tesla market share?

 

  1. Porsche is an electric sports car starting at $90K – at that price point it is competitive with model S not model 3. In competing with the S it comes down to whether one prefers a sports car to a sedan. I have owned a Porsche in the past and would only consider it if I wanted a sports car with limited seating capacity (but very cool). I loved my Porsche but decided to switch to sedans going forward. Since then I’ve owned only sedans for the past 10+ years. It also appears that early production is almost a year away, so it is unlikely to be competitive for 2019.
  2. Audi is at price points that do compete with the Model 3 and expects to start delivering cars in March. However, I think that is mainly in Europe where Tesla is an emerging brand so it might not impact them at all. When I look at the Audi models I don’t think they will appeal to Tesla buyers as they are very old-line designs (I would call them ugly). The range of the cars on a charge is not yet official but seems likely to be much lower than Tesla which has a big lead in battery technology.
  3. The Jaguar competes with the Tesla Model X but while cheaper, appears a weak competitor.

 

I don’t want to dismiss the fact that traditional players will be introducing a large number of electronic vehicles. The question really is whether the market size for electric cars is a fixed portion of all cars or whether it will become a much larger part of the entire market over time. I would compare this to fears that analysts had when Lotus and Wordperfect created Windows versions. They felt that Microsoft would lose share of windows spreadsheets and word processors. I agreed but pointed out that Windows was 10% of the entire market for spreadsheets, so having a 90% share gave Microsoft 9% of the overall spreadsheet market. I also predicted Microsoft would have over a 45% share when Windows was 100% of the market. So, while this would decrease Microsoft’s share of Windows spreadsheets, it would grow its total share of the market by 5X Of course we all were proven wrong as Microsoft eventually reached over 90% of the entire market.

For Tesla, the question becomes whether these rivals are helping accelerate the share electric cars will have of the overall market, rather than eroding Tesla volumes. I’m thinking that it’s the former, and that Tesla will have a great volume year in 2019 and that its biggest competitive issue will be whether the Model 3 is so strong that it will get people to buy it over the Model S. Of course, I could be wrong, but believe the odds favor Tesla in 2019, especially the first half of the year where the competitors are not that strong.

Soundbyte II: The NYC / Amazon Deal Collapses

I never cease to be amazed at how little regard some Politicians have for facts. I should likely not have been surprised by the furor created over Amazon locating a major facility in New York City. I thought the $44 billion or more in benefits to the City and State and massive job creation were such a win that no one would contest it. Instead, the dialog centered around the $ 3 billion in tax benefits to Amazon. All but 1/6 of the benefits (which was cash from the state) were based on existing laws and amounted to a reduction of future taxes rather than upfront cash. What a loss for the City.