What do we have to believe from here?
A bottom-up approach to a top-down problem of market valuations
We recently wrote a memo to TDM clients and it has been re-printed here for public consumption. The memo was a combination of the hard work of Andy Simon, one of TDM’s longest serving members of the Investment Team, and Tom Cowan, a TDM co-founder. You can find Andy and Tom on LinkedIn
Since starting TDM almost 16 years ago, there have only been four times when the markets macro conditions have necessitated communicating our overall view about markets to our clients— two of those times will now be within 12 months of each other (our last note in April 2020 can be found here).
As you all know, we don’t like talking about the overall market and valuations therein. We obsess every day about our businesses, how the businesses are performing and how we can help them achieve better outcomes. As we often discuss at our annual reviews with you, if our businesses perform in line with our 5 year expectations from a revenue and earnings perspective, then we have confidence that we will compound our capital at north of 25% per annum, the hurdle we have cleared now for 16 years. With this business owner mindset, we are fortunate we don’t need to worry about daily fluctuations in share prices — if you were to come into our office, there is only one person who has any idea what the market is doing on given day and that is our wonderful head of trading, Anna!
With everything going on in the world and with markets at record highs, we felt the need to help clients understand how we are currently thinking about the portfolio and the overall markets.
Some of us are now old enough to say we were investing at other strange times, such as:
· The tech bubble bursting in 1999 / 2000 — the Nasdaq fell approx. 80% from peak to trough and took 15 years to get back past the previous peak;
· September 11 terrorist attacks — most indexes fell approx. 15% over 5 trading days;
· Global financial crisis of 2007–2009 — from the peak in October 2007, the major US indexes fell over 55% to the trough in March 2009 before taking another 3 to 5 years to get back to previous highs;
· European debt crisis of 2011 — major US indexes fell between 19% and 30% over a 6 month period;
· Tech meltdown of March 2016 — the overall Nasdaq fell c20% in 2 months, but we can remember clearly the small to mid cap technology businesses that we follow closely generally fell c50% over the same period (as an aside — still to this day, we can’t work out what the catalyst for this decline was!);
· Covid in March 2020 — and of course, more recently, the 30% to 40% fall in 33 days as the world went into lockdown.
We think the current market environment in our investable universe of growth companies— let’s call it the “mid-pandemic world of cheap money, and crazy valuations” will also end up on this list in time … we just don’t know when. And yes, we have the right to rename this period once we know more!
When we think back through the different periods, we would probably liken the current period to 1999–2000 more than any other. We are certainly not at the extremities of what we experienced in 2000 when many technology businesses had very large valuations but close to no revenue. It seemed insane at the time and history has proven this to be true. Conversely, this time around the most highly valued businesses have amazing and easy to understand business models, and large and very fast-growing revenue bases. Because of this, you can have some conviction as to what the business looks like in the medium term — but the point of our comparison is that what people are willing to pay for these businesses is at levels not seen since 2000.
There is a growing sentiment amongst a lot of professional investors that the only thing that matters in a low interest rate, low economic growth environment is the company’s growth rate and that valuations are a secondary thought — we have heard a number of professional managers look to argue that what price you pay actually doesn’t matter in these amazing, fast growing businesses.
To help show how SaaS valuations have changed over the last 10 years, we have charted below global SaaS businesses using Enterprise Value (EV) / Next Twelve Months (NTM) Revenue multiple broken into cohorts based on valuation multiple bands
Some interesting colour in the valuation expansion discussion:
- Finding a business trading on 15x revenue used to be rare, now more than 55% of all these businesses trade on more than 15x revenue.
- The average multiple from 2010 to 2015 for all SaaS businesses on this list is just 5.5x, and 7.0x from 2010 to 2020. This is compared to today’s average of 20x.
We are also seeing a variety of behaviours reminiscent of those from 1999 / 2000 that we were not sure we would ever see again — it seems there has become an accepted normalcy for IPO’s to double from their issue price on day 1, January trading volumes in the US is up 90% year on year (Gamestop anyone?), option trading in 2020 was up 50% year over year, margin lending in the US is at a level not seen since 2008, and of course, how can we not include the $73 billion raised by blank cheque (SPAC) companies in 2020. The famous tautology “it is like Deja vu all over again” is ringing in our ears. Oh, and we almost forgot… last week’s global equity inflows were US$58bn. This compares to the last 10 years where the inflows generally have ranged from negative $10bn to positive $10bn.
So how are we are thinking about our TDM portfolio today in this environment?
We are always drawn to the mental model in times like these that likens investing and portfolio management to that of a great football team manager who understands the importance of balancing ‘offense’ with ‘defence’. This is particularly pertinent to us — there are only ever finite spots (15 companies) available in our portfolio.
We see a key part of our role as capital allocators as being able to know when to play offense and when to play defence. Right now, we are playing defence, with a strong preference for a larger cash weighting in the portfolio over deploying capital. This mental model is not new. It has often been spoken about by revered investor Howard Marks (of Oaktree Capital fame).
“ … we want to have a lot of offense when prices are low, psychology is depressed and the outlook is bad to most people but we don’t think it’s so bad. And we want to have defence when prices are high, psychology is buoyant and everybody sees a brilliant future, and we don’t think the future may live up.”
The idea that higher valuations necessitate more heroic assumptions (or lower future return expectations) to justify holding a position is an investing truism. We always like to focus our attention on what any investor needs to believe to generate an adequate return for holding an investment from here, with the key question being, “if I bought today, what future returns can I expect?”. We strongly believe you can’t just own an amazing business regardless of valuation.
We idolised Warren Buffett in our teenage years — his philosophy around backing outstanding management teams over the very long term has been the foundation of our investment philosophy. However, there has always been one thing we don’t agree with — that you should never sell your shares in great businesses.
A few examples to help explain our perspective:
The Coca Cola Company, one of Buffets most famous investments …
A decade after his post 1987 market crash purchase, Coke was trading at $35, compared with c$50 today. Buffett made 10x his money in the first ten years of the investment (or 25% p.a) but only 16x his money (including dividends) over a 30 year hold (or the equivalent of 10%p.a). He still owns shares. The opportunity cost of holding through the last two decades has been extreme considering the high rates of return he has proven to achieve over long periods.
When we dream of long growth duration, high quality businesses it is hard not to think of Microsoft or Walmart. Since they listed in 1986 and 1970, respectively, they have both compounded shareholder returns at 25% and 20% p.a. Both these businesses though are also great examples of what can happen if your future growth is already baked into your valuation and, despite near perfect execution, an incoming investor doesn’t make an adequate return.
For both Microsoft and Walmart, their share price history can be broken into two distinct periods: from IPO to 2000, and then from 2000 for the next decade and a half.
From listing to 2000, investors in either of these businesses had multi hundred baggers on their hands. What dreams are made of! However, for the next 16 years from 2000 to 2016 both businesses went nowhere from a share price perspective. During this time, both companies grew revenue and earnings strongly — Walmart grew revenue by almost 3x and earnings per share by 4x, while Microsoft grew revenue by over 4x and earnings per share by almost 3x. And yet, if you had held throughout this period as a shareholder, you would have not received a return reflecting this. Valuation matters.
Rather than buying or holding a business regardless of valuation, we as an investment team are continuously assessing and asking ourselves the question — ‘ what does this business need to achieve operationally and financially for us to generate a greater than 20% pa return on our investment over the next 5 to 10 years’ and determine if this hurdle is achievable for management to jump over under a reasonable set of assumptions..
To help explain how we think about this concept and investing at this very moment, we can utilise the Bessemer Venture Partners (“BVP”) Emerging Cloud Comp Index (“the BVP Index”), a publicly available index of 54 US-listed SaaS companies ranging in size from less than $2b to close to $300b in market capitalization. The characteristics of this index help define a “typical” growth technology business today and is a good proxy for our investable universe.
But can we achieve our returns?
The question now is, what do we need to believe to achieve 20% returns over the next ten years for the hypothetical business laid out above? The answer…we would need to be comfortable making some pretty heroic assumptions!
From here, over the next ten years, we need to assume the business can compound revenue at approximately 30% p.a. to more than $7.5b and generate greater than $2b in free cash flow (FCF) . This means growing the revenue base 13x from $550m today and FCF 50x from $40m today. This is reflected in the table below showing expected ten year returns under different revenue growth rates and FCF margins. Our full assumptions are laid out in the end note.
To achieve these assumptions, the resulting hypothetical business places itself in very rarefied air indeed. We can count on one hand the number of software businesses that have achieved this growth feat to date.
The table (figure 2 below) shows the revenue trajectory for a group of the largest and most successful US software businesses over a 10-year period. The starting period, year 0, corresponds to the period when their revenue was approximately in-line with that of the median business in the BVP index of $550m. Of the 9 businesses in this list, just two have achieved a greater than 30% compound annual growth rate over a 10-year period, being Microsoft and Salesforce. Workday, ServiceNow and Palo Alto Networks have not yet reached the 10-year mark, however if they can sustain current growth rates, they are on track to clear the hurdle.
The remainder of the businesses on this list — Adobe, Intuit, Autodesk and Akamai — had strong growth in the early years but experienced a material deceleration at some point in their development and stagnated. In short, to achieve our return hurdle for the median business on the BVP Index today we need to assume performance in-line with five of the most successful software businesses of all time. This is an extremely high bar to clear.
Having said all of that, there is no doubt that the next 10 years will be different from the last and getting to over $7bn of revenue will be achieved by many more businesses than just five (two officially but assuming Workday, ServiceNow and Palo Alto will get there). Digital transformation has been brought forward and there has never been more public high-quality, high-growth technology businesses. In other words, the bar to achieving this type of performance today is lower than it has ever been previously.
The table below shows a group of less mature listed US technology businesses that are on a path to compounding revenue at 30% p.a. over a 10-year period from a c.$500m revenue start.
However, this is already reflected in their valuations, with these businesses trading at a staggering average EV / NTM revenue of close to 40x. While these businesses are growing revenue at an average of 50%+ today, to generate our required 20% p.a. return over the long-term would require them to sustain revenue growth of more than 50% p.a. for the full 10-year period (equating to $40bn of revenue in year 10)! Not even Microsoft has achieved this hurdle with $15b of revenue and a 38% 10-year CAGR over the period. And if we found the next Microsoft, we would hope to earn a lot more than 20% p.a. over a 10-year period!
There is one potential flaw in the above expectations analysis; that is the assumption of maturity in year 10. What if the business is not at maturity in 10 years but still growing at high rates? Surely it would be worth more than 25x free cash flow (equivalent to 7.5x revenue at a 30% FCF margin) at that point in time?
This is indeed a reasonable argument. The problem is, 10 years is a long time by any stretch. Forecasting one year out is hard enough but over 10 years, so many things can happen (positive and negative), that having any degree of certainty over this period is nigh on impossible. While there is most certainly a scenario where the business achieves the high growth hurdle and continues growing at a high rate beyond year 10, there are also many scenarios where this does not happen, and the business ‘hits a wall’ at some point during the 10 years. The key point is, if we must assume the high hurdle scenario to achieve our target returns, there is zero margin of safety in the investment.
And this is where we are today…
There has never been a greater opportunity set of fantastic high growth businesses. But the vast majority of these businesses are already pricing in extremely rosy expectations for the future. This may happen or it may not. We don’t know. What we do know is this — we would much rather invest aggressively when we don’t have to know. Until then, we are happy to be patient and play defence. As is so often the case, we will leave the final word to Warren Buffett:
“be fearful when others are greedy, and greedy when others are fearful.”
As always, thank you for your ongoing support throughout market cycles. Without you, it is not possible to do what we do. Please call if you have any questions.
Tom and Andy
End notes/
The median business in the Bessemer Cloud Index has the following characteristics:
1/ c.$10b enterprise value (“EV”) and c.$11b market capitalization[1]
2/ $550m in revenue over the last twelve months (“LTM”);
3/ 7.5% free cash flow margin, resulting in c.$40m of free cash flow generated over the LTM;
·4/Expected to grow revenue by c.25% of the next twelve months (“NTM”)
·5/Trading at c.15x NTM revenue.[2]
Full assumptions in regards to what hurdle needs to be jumped over to achieve a 20% return from here/
1. We are looking at 10-year returns based on a valuation of the business in year 10;
2. The business reaches maturity in year 10, defined as growing in-line with nominal GDP;
3. At maturity the business will generate a 30% FCF margin;
4. A “through the cycle” valuation multiple for a mature business of this nature is 25x NTM free cash flow (equates to 7.5x NTM revenue at a 30% FCF margin);
5. No dilution from share issuance;
6. Starting cash on balance sheet of $1b (c.10% of market cap);
7. All free cash flow generated is retained in the business.
Short form footnotes referred to//
[2] [1] Compound annual growth rate from year 0 to the final revenue year in the table
[2] As at 20 January 2021. Note that SQ is calculated on a net revenue basis to be more consistent with the revenue of the other businesses in the list
