Invest like it’s 1999
1994 was the year the web exploded. I was a freshman in college. The whole www in September was about 360 web servers. By December, it was over 10,000. Hype surrounding new technologies would show up in glossy tech magazines like Wired, with lots of “revolutionary” rhetoric.
For example, PointCast was a “push based news source”, which was basically a small installable applet that then showed news headlines from major news sources. (Major newspapers were not available on the web yet). A few months after releasing the product, the company was effectively discredited. They created a dud that nobody found useful and that clogged people’s internet connections with useless information. Users already had their news pushed to them in paper form, and they didn’t need to install something to get even more news. But if you had read the news coverage in Wired and most of its competitors beforehand, you’d think PointCast invented the internet equivalent of the printing press. Something didn’t add up.
During the go-go nineties, lots of companies had gone public. Valuations of anything with dot com in the name were sky high. The world was visibly changing. Instead of filing cabinets, offices started using databases. Information became electronic. Hierarchies were leveled. There were pockets of the web, which allowed for immediate purchase of physical goods like CDs, like CDBaby or CDNow, without even picking up a phone. You would even get an automatic email confirming the details of a transaction, within seconds. In the case of amazon.com (AMZN), I remember getting presents like cups with famous quotes for being one of their first loyal customers. While I clearly had the sense that the web was visibly changing many people’s lives, I had trouble understanding the valuations these companies had.
If pressed, most valuation “experts” conceded that it was a fool’s game. A wily Czech friend of mine said, “I’ll keep buying as long as everyone else does, even though I don’t think the stock prices are sustainable at these levels. And then I’ll be quick enough to get out before everyone else.”
Around that time, I started playing with stock picking. While the textbooks I was reading stressed fundamental valuation techniques, that market clearly had different drivers. It was unreal. Was it mass hysteria? One big lie that everyone desperately wanted to believe? Like most of my friends who effectively were retail investors like me, I bought a few internet companies I knew. They did well, not because I had any particular skill. The market as a whole was going up fast-very fast.
I decided to cash out of the market in Jan 2000. I needed cash in order to go back to Poland, for unrelated reasons (i.e. on a complete whim). I decided to liquidate my internet stocks to finance the trip. As a result, I unintentionally ended up being quite lucky with my timing. By the time, I had settled in locally in Warsaw, the NASDAQ and the internet sector started to crash, globally.
That was probably even more confusing than the original sky-high valuations. Companies that recently had completely dominated the stock market disappeared overnight. Investors expected profits of all listed companies. Suddenly investors actually expected companies to be making money, to be earning revenues. It was a harsh wakeup call. Many didn’t survive, although to be honest I’m not surprised pets.com didn’t.
Nevertheless, software (and internet) company valuations intrigued me. First, for reasons other than cash flow, investors were willing to hand over any money they had (probably including their Monopoly© money), frequently based on a few vague promises in press releases. Second, most of these companies had taken massive investments from VCs in the millions, and then gone public. Suddenly they were worth significantly more than their initial value, not having ever generated a dollar of revenue. Then, almost overnight, these same companies were worthless. I had no idea why.
Deeper down the rabbit hole
Out of curiosity, I dug through lots of research reports about the internet sector, both in the US and in Poland. I kept up with the news, even subscribed to news updates from the San Jose Mercury News, to get the latest scoop from Silicon Valley. Even though I was a software developer, I decided to sign up for another degree program, this time a Master’s in Finance. Maybe I was missing more fundamental knowledge. Something was always missing from my big picture. I was looking for real answers.
At the time, accounting professors at the local B-school had a peculiar view of valuation. Major consulting companies influenced them. These though leaders were promoting valuation as a yardstick for strategic management. A whole slew of books promoting fancy acronyms came out, like McKinsey’s book Valuation, going into using Net Operating Profit Less than Tax (NOPLAT) to calculate economic profits generated by a firm. I took a class on Economic Value Added (EVA), which implied that reducing all strategic choices down to the economic value they added. Most of these systems were based on clever interpretations of historically published quarterly reports. Indirectly, strategic decisions affected the stock price. The stock price was the single number summarizing everything a budding executive needed to know.
Yet technology, especially the internet, changes on a monthly basis. Enterprise software, with somewhat slower product cycles, also cannot afford to be backward looking. If senior managers of software companies were supposed to use valuation as a yardstick for making big decisions during the dot com boom, they’d probably all end up in the loony bin.
The truth is: traditional balance sheets are inadequate for reporting internet (or software) company value. A company’s value summarizes how much the company will be expected to make, either through earnings or through capital gains. A software company’s value is off-balance sheet, particularly for startups. The greatest asset a software company has is literally its people, their skills, and their motivation. Other than the number of employees at a company, you won’t find more information in financial reporting of public software companies.
It took one engineer at Google to create Gmail in his free time, a service that was fast, functional, and useful compared to most other available offerings. You can easily scale well-crafted software. On the web, it can affect millions of people. At that scale, you can easily turn it into a revenue and profit opportunity. As a result, traditional analysis of balance sheet ratios and income statements is next to useless, when trying to estimate the likelihood of this happening.
Does this mean centuries’ worth of accounting lore, going all the way back to Fra Lippo Lippi who invented double-entry bookkeeping was inapplicable in the case of software companies? In the nineties, cash was only important in terms of how many months a startup had to live. Revenue was irrelevant. We can be bought out by (big company/dumb money of the day). We were going to build a better way to (lots of techie acronyms no one except geeks understand).
Not really. But what was going on?
It wasn’t until I stumbled into an internship at Credit Suisse in 2004, that I started to get some insight. In equity research, at least the summer traineeship variety, the main focus really is keeping up with the news on specific companies and modifying existing financial models in massive spreadsheets. These models would be used to generate new investment ideas, and to provide an estimate for company value.
The office where we worked had runaway air conditioning. Even during a sweltering July, I had to wear sweaters to the office, and constantly walked around with a light cold. The temperature, however, was an apt reminder of how our analysis should look: cold.
For the sake of anonymity, let’s call my boss Yoda. While Yoda was primarily interested in the oil sector, he also covered two domestic IT companies in Poland. Ah! Finally, I had found a real practitioner of valuation.
Yoda was a rather mousy character, with a massive brain that remembered every number he came across, calculating balance sheet and income statement ratios in his head in real time. Years before, Yoda was in the top 10 of his class when studying at my B-school. He was one of the first CFA charterholders in Poland, at the time a quaint “emerging market” for the more adventurous investor. Investors abroad relied on his expertise whenever thinking of doing business locally. My finance professors looked like complete dilettantes next to him.
Now commanding his full attention, I expected Yoda would give me the unvarnished truth. Having read lots of glossy reports on internet company valuation previously, I could finally pepper one of these report authors in detail with questions, like a mad busboy at an Italian pizza joint.
After coming back from the earning results release of Kompro, the largest domestic systems integrator, we started talking shop.
“What gives, Yoda? Why is Kompro so undervalued?”
“What do you mean?” he said.
“A few years ago, this company was worth almost 3x times as much. It’s not like that much has changed, especially not earnings.”
“Well for one, I don’t trust them”, he confided. “I mean management. They’re staying within what they’re required to report, but I just don’t trust the numbers they publish”.
“Yeah, but that hasn’t really even changed either. This is the biggest local IT company. They generate serious revenues, not vaporware. They are going to keep growing for many years, along with the rest of the economy,” I countered. “And they’re priced pretty poorly.”
“For one, you need to separate out the value of the business from the price of the stock.”
“Huh?”
“ You can have a great company with a really bad stock price. You can also have a terrible company with a really high stock price.”
Like pets.com, I thought. He has a point.
“Hrmph. But then what about all of these internet companies that had no revenues for years? If they aren’t earning anything, why do the stocks have any positive value? For example, Kompro has this little subsidiary, where they put in about one mln PLN ($250k), and now they are earning 24 mln PLN ($ 6 mln USD) annually ever since. That’s a 2400% return in the first year, and millions ever since with no additional money in.”
“That’s just standard DCF valuation. Yeah those are amazing returns, but it’s also a really small part of the company. DCF helps bring everything to the lowest common denominator, i.e. a dollar today, regardless of which business you are in.”
Snooze-regurgitated theory from my accounting classes. That’s the “secret”? I was admittedly disappointed, and kept probing.
“What about these internet startups, that don’t earn any revenues, but VCs are willing to pay millions of dollars to get a small bit of the action? So they’re bad companies with high prices?”
At the time Google was still a privately held company that didn’t sell Adwords yet, and Zuckerberg was just moving his small team of Harvard hackers to the West coast. And there were a lot of recent dot-com casualties plastered all over the news.
“Not necessarily. They can be good companies with high prices. If the investors expect high growth, you can work out what the company’s worth. Try using Gordon’s growth formula with earnings.”
“But these companies don’t even have revenues. Their products are really useful, like Google’s search engine.”
“At any point, they can figure out how to start charging. They have the option to charge. They’re also becoming more popular every day.”
“So having a realistic option of earning revenue is good enough for valuation purposes?”
“Yes. At least for enough people with enough money, so that it makes a difference. Check the options tab on the Kompro spreadsheet.”
Sitting across from him, I look down on my screen and click on the tab, and look up. There is an option chart with time value decay in months.
“It’s not that great of a model, because the options aren’t that liquid, but using a series of options is a useful way to look at these types of companies.”
“So basically the options on the company stock say more about its value than the stock itself?”
“Well, not really. They depend on the current and expected value of the stock. But they do give a picture of what the options market expects the stock to do in the future. And these bettors are willing to risk money on it.”
“But these options are focused on profits, not revenues.”
“True. In an IT company, most of the costs are variable. The cost of capital, employees, hardware, they can easily be ramped up, and in a good IT company costs stay at a really low level relative to revenue anyway, as the revenue hockey sticks up.”
“So costs aren’t that important?”
“They’re always important, but their relative importance is much lower than in a fixed, sunk cost industry like oil, where expanding capacity is very lumpy and much riskier.”
In IT, the revenue side was much more important than costs. The option chains on the stock in this case were the closest financial approximation we were going to get from public information.
He continued, “Of the business functions, marketing and sales are usually more important than finance in IT. Not to mention, you can tell a lot about an industry, based on which function the within the company the CEO typically comes from. Boards tend to promote the guy from the area that will have the biggest impact on profits. In the oil industry, it’s often the CFO. In IT, the head of sales is a more common choice.”
“That makes sense, since the profits can come from either increasing revenues or decreasing costs. But then what about the value of the company? With Kompro, it’s relatively clearer, since they have both earnings and cash flow. What if there are no revenues?”
“We only have public information available, not management accounting level information. So it’s always a best guess,” he parried defensively.
“We know what these companies do. We also know what’s going on in their respective markets. Isn’t that enough to estimate strategic option values?”
“It’s not a great approach, but probably the best we have. And you’re right, that we have much of the same information about their markets available to us, in order to estimate growth rates of various business segments for example. You can use these option values to estimate cash flows with greater precision and frequency, if you want to get into that much detail. The implicit promise of the longer-term strategic option captures most of what you buy, when you buy the stock. You know, the profit drivers,” Yoda summarized.
“So then how do you value these companies, and how does management increase that value?” I said, influenced by the trendy consultants.
“From my point of view, these Internet companies have more in common with media companies than with hardware ones. It’s all about having a product that scales up well, to many eyeballs, and having enough of the right talent to make it all work together. Unlike media companies, these companies are solving specific problems or fulfilling needs, not just providing information. Hardware is a more traditional business, just buying lots of parts, assembling and servicing them, mostly trying to maximize a gross margin. It’s a lot harder to differentiate yourself with hardware, as it’s pretty much a commodity. So an integrator like Kompro is some mix of the above.“
“If you are investing in a such a company, it’s going to be in helping them beef up capacity to serve each market niche where they exist. They need to clearly differentiate themselves if they want be competitive. It’s a brutal industry. Once they know what niches they want to exist in, it’s all about out-executing their competition. Probably the three most important assets are their people’s motivation to serve niches, skills when creating something new, and the internal systems within the company, all of which are not in the financial statements or annual reports. They are all massive implicit call options on future cash flows.”
Hidden hockey sticks, i.e. implicit strategic call options, aren’t reported on balance sheets.
What this means for you
From a business point of view, Yoda gave me a pretty accurate quantitative summary of valuation in the startup world, and probably even the software industry as a whole. The traditional financial reporting framework is at best inadequate, not only because it’s backward looking, but primarily because it doesn’t report what’s really important to understanding software companies anyway.
Given that this is the case, probably the worst thing you can do is commit to long-term financial budgets. It ties your hands, and makes it impossible to respond to customer requests and new market needs effectively. You won’t be able to exercise your call options, if you make detailed plans (and budgets) about next year only using the information you have today. You will be frozen all year in strategic assumptions made on Jan 1. Of course some planning is always a good idea, but you want to commit to as little as possible early on.
While you can expect your weighted average cost of capital (WACC) and your staffing costs to remain to remain relatively stable, you want to focus on big new revenue opportunities with existing products and clients, and also new markets adjacent to your existing habitat, which would be relatively easy to enter on a small scale as a test. If you can tie your marketing efforts with your new product development, you can test as much as you want. Once you have a proven new business, then step on the gas pedal, and get serious financing to milk that proven opportunity.
The really big strategic hockey sticks have high gamma, I can add now based on the CAIA readings. The call options fluctuate in value continuously, even if the value is not tracked anywhere explicitly, based on new information in the competitive environment. Complementary technologies, network diffusion effects, and new tech breakthroughs can quickly change the relative attractiveness of any given strategic option.
Your product backlog is your opportunity set of new product or feature options. Most of your time should be spent evaluating priorities amongst the existing options, getting rid of the useless ones, or adding new ones. You want to find the biggest hockey sticks, and get rid of the puny ones. It’s not that different from hedge fund portfolio management, just with real options instead of financial ones.
Basically, think agile-even in the finance department.
At least Yoda didn’t use that cheesy line “Use the force, Luke”.
[Disclaimer: some names and identities in this post have been not so cleverly disguised]
Related articles
- Intrinsic Valuation…the Basic Framework (stockvaluationcanada.wordpress.com)
- A billion dollar software company is founded every 3 months (minming.net)
Filed under: Software Tagged: NOPLAT, PointCast, valuation
