🔒 Cavalcade of unicorns going public – where should you invest?

LONDON — It looks as if 2019 will be the year when a number of high profile tech companies are going public. Ride-hailing company Lyft is first in line, followed by its bigger rival Uber. They are not the only companies who have signalled their intention to float; messaging start-up Slack, image company Pinterest and food delivery company Postmates are all moving closer to IPOs. There is good intentions in the flotations of Uber and Lyft for their drivers, as both companies have indicated that they are planning to offer cash bonuses to some of their longest-serving or most-active drivers which they can put into shares in the company. Estimations for the value of  two ride-haling companies have been estimate at between $20bn and $30bn for Lyft while Uber comes in at around $120bn. In an interview with Bloomberg’s Joe Weisenthal, Rett Wallace of Triton.ai gave a list of factors that investors should consider before investing in newly-listed IPOs. He says big IPO listings have become less common. – Linda van Tilburg

Like most things, there are a couple of different narratives by way of explanation. What you only hear from people in Silicon Valley is that founders don’t like to take their companies public because being a company’s CEO is kind of a pain in the neck. So, anybody who’s having to comply with the SEC and the Act and other things that were imposed on publicly traded companies as a protection against retail investors who buy their shares.

If you have access to capital in the private markets it might be easier for you to stay private and not expose your numbers, expose yourself to liability and so forth. So, almost all of the reasons that you could think of, why a company wouldn’t go public are regulatory. They stem from the changes and the regulations of the securities businesses and there’s a long key narrative that we could get into about that.

When companies like Amazon, Netscape, and Yahoo went public on a couple of million dollars of sales and earlier generations of companies, like Intel, and so forth went public. Really, as soon as they got into revenue that’s because capital formation happened in the public market.

Let’s backup for a second. Tell me about your firm and why this is an area that you pursued? What is it about IPOs that are interesting in general, and what is your background that caused or prompted you to go into analysing them and providing this surface of breaking them down?

Sure, well IPOs are a very good example of what came after the ‘Great Depression.’ When the government decided to reform the securities industry so that you didn’t have a big, speculative bubble anymore, the kind that created the 1929 stock market crash. So, the innovation at that moment was that securities come with data. Stapling the 10-Ks and the 10-Qs, the regularly recurring reporting and disclosure so investors would know what they are buying was the great innovation after 1929, and that prompted the situation where companies, if they wanted to trade stocks with each other or have people buy their shares, they had to be public. So, companies went public much earlier. If you, again geekily read the biography of Rockefeller, for example. Before the crash, one of the reasons he was so successful in investing is that he had access to information that wasn’t broadly available.

That always helps.

Right. So, information has always been a key component of being successful as an investor so, the origin story of our firm is that we saw what was happening that fewer and fewer companies were going public. That meant that more and more of the interesting companies were private and these companies operated outside of the information regime of the securities acts of the USA.

The other thing that we noticed is that all of the information architecture that was installed as the operating system of the securities trading institutions was developed in the 1930s. So, generally, accepted accounting principles – some people will tell you it’s like the perfect information that you could have about a company but it never existed until 1938. Like Moses did not come down from the mountain with Gap. Company categorisation, the standard, industrial classification system again, like the 1930s and so, these pieces of data architecture haven’t iterated and advanced. So, we’re still sort of stuck in the 30s with the way companies are analysed. So, the origin story of our company was, we were looking for ways to be smart about investing in companies that were generally private companies, and the architecture that people used to look at public companies wasn’t particularly serviceable to that end so, we had to build a new one.

So, obviously when a company files to go public and it files its S1 to the SEC the company engages in the practice of putting its numbers into a type of a structure that’s similar to other public companies or identical.

Correct, there’s a template that everyone had to adhere to.

But there’s still the problem of investors hasn’t really gotten to know these companies and even within generally accepted accounting principles, there’s all kinds of idiosyncrasies and opinions and different approaches and companies that have been public for a while people become familiar with aspects of their business model and they understand the moving parts, and that doesn’t just doesn’t exist yet, certainly at the time of the S1 filing. So, when you look at an S1 filing, besides the obvious, the balance sheet, the income statement and the cashflow statement, what else are you looking for when you start to breakdown and looking at these companies from the perspective of an investor?

So, our point of view on companies is that a company is really just a receptacle for different product lines. So, for example is that Uber Apps and Uber Eats live inside the same company but they’re totally different businesses and completely different product lines. So, as companies goes public much later in their life, what it means is that the audit of the consolidated entity disguises all of the individual operations that are happening inside of a company that might have a bike sharing business and a scooter sharing business and operates all over the world in different types of jurisdictions. So, the bigger it is, the harder it is to get your arms around it unless you can see the detail.

So, that’s really an interesting point. So, if a company is just in the business of making widgets then you can have some sense of like, okay widgets cost the company this much to build and raw materials costs this much and labour costs this much, and you sell the widgets for this much and then you’ll look at the gap between cost and the sale and you know something about the business.


But with these big companies and with new businesses that people don’t understand, with novel business models, simply subtracting cost from revenues – it just doesn’t tell you that much about the company.

Yes, the architecture of a digital company is just completely different than the architecture of a 1930s railroad or a metals and mining company. One of the things that again, geeks, that have spent studying a lot of time studying how Gap works and have suffered through accounting class – inventory accounting is one of the things that is really painful in the FIFO and LIFO kind of stuff. How do you track the inventory of Facebook?

Well, so then that gets to the question, going back to the Uber example. Obviously, it’s still primarily a car sharing company but in many different businesses, and they do also have several different lines, and in some places, they have scooters. So, how do you go about essentially trying to disassemble the business from these consolidated financial statements?

So, when we started out, we were looking for ways to be smart about how to tell which dog-walking app was going to be better than the other dog-walking apps, for example, because you listen to the young entrepreneurs pitch you a company and it always sounds good but you don’t have a comparative base of data. So, the SIC code system was no use to us whatsoever, in how to categorise companies into the bucket of dog-walking apps, and then figure out which one was going to be the best dog-walking app. So, we had to design an architecture that you get the apples and apples in the same buckets and separate them from the oranges and the grab apples, and the tangerines and everything else.

Read also: The ripples of disruption: Uber blows a hole in airport finances

So, one of the things that was fairly funny about this is, if you use a sort of top-down SIC level type categorisation system and you use a word like transportation – what we found is that companies like Uber bucketed into the same bucket as Zipcar, right. But you look at, and you’re like okay, well, Uber doesn’t own any cars. Zipcar car owns thousands of cars that they have to park, maintain, fuel, paint and all that sort of stuff. So, it’s like okay, even though from a narrative perspective these things look the same. There’re really not the same so, our response to this was to flip everything upside down and to look at how the thing works in terms of what does the customer pay for and what does the customer actually get. In this example, if you’re trying to go to Brooklyn from Manhattan you could rent a car with Zipcar and drive it yourself, or you could have Uber drive you there and it just turns out that the mechanic of the system that delivers a ride versus the access to a car are totally different things.

Right, so is there enough information straight from the S1s, or I guess Zipcar has been public for a while, right, to actually perform that calculation or do you need to go elsewhere?

What’s great about it is usually you don’t need the S1 to know how a Zipcar works because Zipcar tells you everything about how it works on their website. So, if you flip the thing upside down and look at it like a user, it’s actually not very difficult to figure out how these mousetraps work

Now, one of the things we’ve talked about, because we’ve talked on our on-air on TV before is non-financial statement characteristics of companies. So, people are interested in things like just the level of transparency period, structural, things like voting control. What are the other things that you look at when you analyse a private company or soon to be a public company, beyond just the numbers?

Sure, one of the things about Gap is that Gap translates everything into dollars. So, the numbers you see on a Gap PNL are all dollar denominated. But most of the numbers that are the most interesting about companies aren’t dollar denominated, like how many customers and how much do they pay and how long do they stick around and where do I get them from, and things like that?

Or just how many cars they might have an inventory of.

For example, right, and so there’s a big debate that you could read about Matt Levine here is very articulate on the subject about non-Gap reporting, and some people get kind of religious about this and say that you shouldn’t report things that aren’t Gap because those companies aren’t comparable anymore. But the problem is that if you only have the PNL, like for example, if you were looking at the Snap IPO and you saw that Snap lost $1bn in the trailing year, you don’t know very much about Snap but the intuition that people have about that company is well, I know my teenager can’t put it down but you don’t have the statement about how many teenagers and how long they stick around, and what you definitely don’t have is the statement of how advertisers and how long they stick around and how many sales people it takes to get those advertisers to pay you and so forth. So, to us, again, the numbers that matter are the numbers that help you calculate the mechanics of how the mousetrap works and those things are often not disclosed, if not at all and you need other ways to go get them.

What do you think…? It’s interesting you mentioned Snap and maybe this is a slight tangent or maybe not, but it feels like there have been efforts with a lot of these internet companies to essentially standardise some of these non-financial metrics. So, MAUs (monthly average users) is popular way to compare them but it feels like the companies are really pushing back against that, or like to, and they want to create their own bespoke ones, and they say, ‘no, you can’t compare our MAUs to Facebook or RDAUs.’

Community adjusted EBIDTA is my favourite one of those.

Yes, Twitter recently announced that they were going to, for the first time, start revealing DAUs (daily average users). They’re no longer going to report monthly average users but even their DAU numbers – they’re calling them MDAUs (monetizable daily average users) to distinguish from users who they’re probably not going to make any money from, so their MDAUs aren’t going up. Anyway, the point is, what is your view on this? Do companies have an incentive to try to breakout of the standardised comparable numbers and come up with their own vanity metrics that are always going up into the right?

I think the world doesn’t abide on this. Like people don’t like accountability right. So, if you don’t have to be accountable to a particular metrics you’d rather not. One of the things that’s interesting about what’s happened in capital formation right now is that private company investors have access to all this information, all the real information. Not the fake, monetizable daily average users – they can see all of that sort of stuff and they have a real sense of how those mechanics work. Once you arrive in public company land many of those numbers are not disclosed and mostly you find a situation where as capital is forming around these companies the investors that put up the money have much better access to information so, more transparent situation but in liquid situation. Then you trade liquidity for transparency, in the sense that the public investors don’t really get to learn any of way that the mousetrap works but at least they can sell the stock, and so that’s the trade.

Read also: Big tech takes on anti-vaxxers – The Wall Street Journal

As far as your question about the standardisation, and sorry to go on so long, using like an ad supported company’s metrics to analyse the subscription business is just not very helpful. So, engagement metrics, for example, people ask us about our engagement metrics, which I always laugh because I think engagement is bad. We want our users to figure out the answer in as little time as possible because I’m not trying to serve an ad to them, right. So, each company is different and this is what we’ve spent years doing is developing an architecture so that you understand what kind of company you’re looking at, and look at the appropriate metrics to do so.

Going back to what you were saying about the trade between liquidity and transparency. We had a recent episode a few months ago. We were talking to a VC and he was arguing that one of the things that made this period in market unique is that, whereas in the past illiquidity was a penalty for a company, and a private company had to offer a bigger premium to get private capital because that was more locked in. These days, people are paying a premium for access, in his view, to illiquid companies. Maybe they didn’t want to have to mark their books day-to-day or maybe there was some sort of prestige value of being in a Lift or an Uber that caused people to overpay. Do you see that that these sorts of traditional discounts that would have in the past, come along with private equity stock has flipped?

I’m going to give you two answers to that question. One, just to verify with data, the claim that people pay a premium.


Over the last five-years the IPOs that we’ve looked at tend to trade-up in the first-half of their first year that they’re public, then in general, trade-down again below the IPO price. So, public market investors…

That’s over the last five-years so, we’re not just looking at the 2018 affect here?

Correct. No, and it’s not just Alibaba or whatever. It’s 150-odd transactions so, what happens is, as a capital markets matter these things come out. The IPO prices as it begins to trade, you get the famous ‘POP’ which some people love and some people hate, and in general, these things trade-up for a while and then large amounts of shares are unlocked and people and people start paying attention, and they change the channel, and they look at something else, and then they trade-down.

So, it’s definitely true that private market investors have paid a premium. Like the guys who bought the last round of those deals could end up under water if they didn’t sell but they could sell. So, it’s unclear if they’re being penalised for paying that premium because they had a moment where they probably could have made a profit on the trade. But in general, it’s not a good trade for the broad-base of shareholders so, that’s item one.

Item 2 – Why do people pay a premium for this? The answer, we think, is because if you want to invest in growth companies you have to pay the price. So, if you are a long only manager, who’s managing a growth fund, who has a carve-out that allows you to invest in Uber Lift, or whatever.  Your ability to set price is very limited but you want to participate in those deals and as more and more capital has flowed into this space – what happens when there’s more demand than supply…? Prices go up.

So, it’s kind of like… It’s a function of the fact that even if we were just looking at public markets, we know that the growth factor has done extremely well in recent years, and that’s just even more exacerbated in the ultra-high growth private area. So, that could explain, at least, part of this premium.

Sure, if you were a growth investor in the 1990s, you would be investing in companies publicly that were young, and you’d be buying Amazon, Yahoo, or the Globe.com – you buy the good and the bad, but you get to do all of that on the public market. Now, all of that capital formation and all of that value appreciation happens in the private market and the guys with large pools of capital want to participate in that but what it’ also done is, it’s created a situation where larger pools of capital, the vision fund, for example, have formed to participate in that trade.

I’m glad you mentioned the IPOs prior to the bubble because obviously, everyone knows you’d be rich if you had bought into that Amazon IPO but you would have lost all your money if you had bought into the Globe.com IPO. People bemoan the decline of IPOs precisely because they have memories of Amazon and Microsoft in their mind and they say, the stock market used to be this avenue where people could make a lot of money, investing in these companies Now, that’s closed off to anyone who doesn’t have access. But of course, it does seem like there’s a lot of hindsight bias because most companies are more like the globe, right.

Sure, Mary Meeker has a great statistic that was in her deck for a long time, after the bust. That 2% of the companies that went public during that moment in our culture created more than 100% of the returns.

So, the vast majority of them were total flops?

Totally. If you were in the 2% that compensated for the negative j-curve, right so, the vast majority were a bust and more money was lost than made in aggregate so, you had to be very picky to not be one of the losers.

Is the decline of the IPO a bad thing?

It depends. If you were a retail investor and in hindsight, you’re totally convinced that you would have absolutely put your life savings into Uber, if you had been able to buy it five-years ago, then it’s a bad thing. But one of the reasons that the bar has been raised so much for companies to go public is to protect retail investors from themselves. Retail investors fuelled a lot of the bubble that happened. There are other structural reasons why the internet bubble happened but there was a huge amount of demand in the same way that people now speculate in cryptocurrencies and other things like that because it was perceived to be an easy buck.

Yes, people are always going to look for action somewhere. So, as of this moment, while we’re recording, and we don’t know. Any day now, we could get S1 filings from some of these companies that we’ve mentioned. So, Uber, Lyft, and Slack and a bunch of others that we could get, for the first time, public data on these companies. So, when these come out what are going to be the first things that you look at and what should people listening at home – what should they start to look at specifically?

Well, us first. Our system is just to take it apart and do the 15-point inspection on these things. So, does the math make sense? Does this company make money, is one of the things that we’ve talked about on TV before? There are times where we’ll put companies into our model machine and we’ll look at and see, geez, there’s no setting of the model that produces a profit, ever. So, that’s a really low score, as far as the earnings power of the company. But we also look at the management team. We look at the founder power.

When you say, you look at the management team…

We score them.

How do you score, in theory, the quality of a management team?

The things that you think that you would think to do if you wrote out a rigorous system. Have they done it before? How long have they worked together? Have they worked in places that you’ve heard of before? Were they successful there? Did they go to schools that you’ve heard of before? Do they have advanced degrees? Then, when you toggle to the founder aspect of the management team. Sometimes you see total control of the founders, which tends to be great because they’re highly invested and have a lot of skin in the game. Sometimes, like for example a Twitter you see the company is totally post-founder and that means that the management team has economics that are heavily weighted towards the upside but doesn’t have a lot of pain associated with the downside so, founder power is very important. The quality of their board. The quality of their investors is interesting. How famous is it? Like, fame and buzz, is one of the things that we score our companies that nobody has ever heard of, who do less well, than companies that are well-known.

Okay, so you have all these factors, 15 different…

Yes, 15 different scores that roll up into the summary score.

So, 15 different scores so, in your experience, the aggregate higher scoring companies do better than the lower ones.

Way better.

Otherwise you wouldn’t have a business.

Totally right, but shockingly because there are times where we get the score, because we see it when it comes out of the machine and we look at it and we’re like, ‘that can’t be right.’ So, it’s always interesting to us. Like, maybe this will be the one that we have to rebuild the whole system on.

Here’s my final or the key question I have is, do high scorers say, you should invest in this company or is it if you invested in every company with high scores and shorted or avoided all the companies with low scores – would that be a superior strategy?

Yes, so the aggregate trade is always better, unless you are so good that you can sniper shot the singular winner but that’s incredibly hard to do.

But the premise of the scoring system is essentially that on aggregate, you’ll strip out a lot of noise and be much more likely to have a winning portfolio of IPOs, with the higher scoring companies?

Not only that but that’s certainly true if you’re an institutional investor. Most of our customers are institutions that buy at the IPO price so, the returns are three-times better if you buy the high scorers than the low scorers. But if you buy the first trade, if you’re a retail investor, buying high scores versus low scores – this is the difference between making money and losing money.

Visited 40 times, 1 visit(s) today