Financial engineering is a powerful tool. Here, we see it transform negative-sum risk into positive-sum fun.
Nobody knows what advice is worth, which is why there are three different ways to charge for it, none of which really make sense.
The cheapest tier of advice is the kind you get for the cost of negative-one cups of coffee. You email someone who has been doing whatever you do, or want to do (you will inevitably refer to this as “reaching out”); you suggest coffee; they buy you coffee and do some combination of telling you what to do better and dissuading you from trying. The secret to getting this kind of advice is knowing who to ask: anyone famous will almost certainly ignore you, and anyone who isn’t famous will be incredibly flattered that you think they’re worth talking to. As long as you’re not asking for “advice” but actually trying to sell something, it will probably be the highlight of their day.
The middle tier of advice is provided by consultants, lawyers, accountants, and investment bankers. The price ranges from $5,000 to $50 million, and there’s flexibility at either end. If you think about what the phrase “the business of giving advice” means, this is probably what you think of. Someone has a specific problem; they hire someone who has seen that problem before; problem solved. (Hopefully.)
This tier is the least generative of insights. The business is mature. Buyers and sellers are savvy. It’s lucrative enough that when consultants burn out, they can retire and write books. And while the books aren’t tell-all books (one assumes the big white-collar service firms write some tight NDAs), they’re tell-some books, and they tell us about what we’d otherwise guess: that these companies are in the business of helping good ideas (or, if you want to bill more per hour, “best practices”) percolate through industries. This is socially useful, especially compared to the alternative of industrial espionage.
The top tier of advice is sold completely backwards: you have to convince them to take you on as a client, and they give you money. They take some equity in exchange. The term for this subset of the advice industry is venture capital.
Superficially, venture capitalists are not in the advice industry at all. “Capital” is right there in the name, and “Venturing” is the part they’re not involved in — you’re the one who has to take Caltrain all the way to Menlo Park to pitch them. It’s a misnomer that dates back to the 70s-through-mid-2000s, when there was a shortage of money and a surplus of ideas, so financiers had the power. Since then, for a variety of reasons, things have reversed: startups have a lot of negotiating power, especially if all they want is money, so venture capitalists have to offer more. And what VCs have morphed into is advisors.
Of all the people who need advice, startup founders should be at the top of the list: they’re necessarily inexperienced at whatever they’re doing, given that a) this is usually their first time in charge of a company, and b) that company didn’t exist before. As companies scale, everything breaks — it’s just math, but scary to point out, that in a company that trebled its headcount in the last year, the median employee has been there about eight months, so every process that isn’t well-defined has been subject to a brutal game of telephone. Meanwhile, growing companies face new challenges, like competitors gunning for them or trying to buy them (or, if they’re playing hardball, doing both), breaking laws they didn’t know existed, finding the parts of the business that get more than 2x harder when you’re 2x bigger, etc.
Basically the only people who have anything useful to say about the specifics of this, and the psychology of it, are former founders, and it’s no coincidence that former founders are overrepresented among VCs and extra-overrepresented among the best VCs. It’s a fortunate economic coincidence that when someone sells a company, four things are true: 1) They have a high liquid net worth, 2) They have a lot of accumulated wisdom about running a company, 3) they have a lot of friends, including a lot of new friends in the lubricating-a-liquidity-event businesss, and 4) they are very interested in not running another startup for a long time. Investing is a good way for them to stay in the game, but when they get the dreaded 2am phone call, they can ignore it and go back to sleep.
Advice to startups faces different incentives than the advice to other entities, because the results are so variable. If a consultant tells a steel mill how to revamp their shift structure, that might raise productivity +/-1%. But a venture capitalist advising a founder on whether or not to accept an early acqusition offer is pushing a decision where the results vary by an order of magnitude. When the outcomes vary that much based on your input, the only compensation scheme that makes sense is equity. Equity aligns incentives, more or less: make the company twice as valuable, you get paid twice as much.
But that “more or less” hides some complexity; often equity is not quite equity — and importantly, different ways of structuring equity-like deals provide a way to bridge the gap in information and incentives between funders and founders. Fitting a key to the misaligned-information-and-incentives lock is generally known as financial engineering.
Financial Engineering Makes Us Better-Off
“Financial engineering” was originally a pejorative term, but it’s getting less so. This isn’t because people reclaimed it the way other pejoratives get reclaimed; it’s just that the kind of people who are really interested in financial engineering are also really socially unaware and we don’t notice when you’re making fun of us.
It refers to any business activity that involves no change to the underlying business, but significant changes to who is entitled to which cash flows. Borrowing money to buy back stock? Financial engineering. Selling stock to pay down debt? Also financial engineering. Issuing weird convertible debt and using it to buy back the warrants you issued as part of the deal where you swapped your preferred stock for zero-coupon debt? Obviously and definitely financial engineering.
Hollywood is full of creative financial engineering, because every movie is basically its own temporary company, and a lot of the talent employ full-time negotiators. So, sometimes you get some striking deals, the funniest of which is that Arnold Schwarzenneger’s most personally lucrative movie was the buddy comedy Twins:
[W]e sat around at a restaurant and we made a deal on a napkin and wrote down this is what we do. We’re going to make the movie for free. We don’t want to get any salaries and we get a big backend. Ivan should take this deal to the studio. He’ll give it to Tom Pollack who was then running Universal Studios Tom Pollack said this is great and we can make this movie for $16.5M if you guys don’t take a salary and you get a big backend. We’re going to give you 37% or whatever for Danny, Ivan, and me. We worked out the percentage of what our salaries are. Whatever Danny got for a movie versus what I got for a movie versus what Ivan got for directing, we worked out percentage wise and that’s how we ended up dividing up the pot amongst ourselves. Let me tell you, I made more money on that movie than any other movie and the gift keeps on giving… [Pollack] said he’d never make that deal again.
Financial engineering matters more late in a company’s life, when it becomes possible to predict the size, growth, and variance of their cash flow with a high degree of precision. Once that becomes possible, lenders get interested in making loans backed by that cash flow — and when lenders start offering better rates, borrowing money to cash out stockholders becomes a sensible trade. Meanwhile, the company’s business can be split into different cash flow streams with different variabilities. Across several industries, businesses have split between sort of abstract companies that own and manage brands, and specific companies that own hard assets in the same industry.
As a rule, you’d expect that lower variance in outcomes means more rewards from cleverly-structured debt, while higher variance in outcomes means more rewards from cleverly-structured equity. However, variance in outcomes tends to be higher when a company is young, and it would be silly to spend 10% of your post-money valuation on fees to the banker who structured the transaction.
In fact, it’s that very complexity that seems to have killed a useful financial tool a few years ago, and now is the time to bring that tool back: high liquidation multiples for early-stage preferred stock. The two most likely reactions to this proposal are yawns and pitchforks, but as it turns out this is an elegant solution to a common problem.
But first: what is it?
A quick tour of the cap table: common stock is what you think of when you think “stock.” The shares you own represent a proportionate share of the business. Preferred stock is common stock with some extra features, the most important of which is the liquidation preference, which by convention is 1x. A liquidation preference means that if you invested in a company and they liquidate, you get back 100% of your investment before the common shareholders get paid. So if you invest $1m in 1x preferred shares in a company at a $10m post-money valuation, you get either a) 10% of the value of the company if it sells for over $10m, b) $1m, if it sells for between $1m and $10m, or c) Whatever it sells for, if it sells for less. With a 2x liquidation preference, the same investor writing a $1m check on otherwise identical terms would have the same economics for a big win, but they’d get the first $2m instead of $1m. All else is not equal, however. Since the investor has more protections, their investment should represent a smaller percentage of the company.
Different parties involved in a startup have different risk tolerances: in general, anyone who sells the company a one-time service or a commodity product will want cash, and everybody else will want equity roughly in proportion to how much they could singlehandedly kill/save the company, so: lots for engineers, product managers, and senior managers across the system, less for others. There are exceptions — one guy got Facebook stock at the A round valuation in exchange for drawing a grafitti mural on the wall of their office, and Paypal’s landlord took options. But those situations almost exist as a sort of countersignaling opportunity, as if to say “We know this looks like peak-of-the-bubble behavior, and, you know what? We’re going to do it anyway, because we can.”
Venture capitalists’ risk tolerance is hard to describe. In one sense, they have the most out of anybody involved in an early-stage startup, since they’re the most diversified investors in the company. In another sense, though, they’re fiduciaries, and they’re human beings who don’t want to look stupid, so they’ll often structure deals to be pretty simple at the early stage (when everything usually goes to zero), and a little safer at later stages, when things can go wrong-but-not-catastrophic.
Liquidation preferences, sometimes elaborate, are more common in later stages. They give investors a bit more comfort about otherwise stretched valuations, and they offer investees a form of bragging rights, too: a company’s purported valuation sounds higher if you just take the value of the preferred stock and divide it by the percentage of the company it represents, even though this overstates the valuation; in effect, the liquidation preference is an at-the-money put option with an indefinite expiration date on an extremely volatile asset, and those aren’t exactly free.
It used to be more common to have liquidation preferences of 2x, or even (I’m told) up to 10x during the 90s and early 2000s. These deal terms got a bad reputation, in part for good reason: if you’re a VC who has seen lots of term sheets, and the person you’re negotiating with has seen only one in their entire life, you can instantly make yourself a lot richer by saying “Oh, that? Yeah, it’s standard.” Because the people who used these terms in that way tended to undercut their investees in other ways, too, founders noticed a correlation.
But while there’s a dishonest way to play fast-and-loose with terms, the terms exist for a reason. High liquidation preferences are useful because they realign incentives, and in particular because they help solve one of the biggest conflicts between venture capitalists and their investees: VCs round every exit under a billion dollars down to zero. That’s stylized, but only slightly: the venture capital business is hugely dependent on outliers.
A completely honest discussion between a founder and their board that illustrates this dynamic might go like this:
Founder: “We have an offer to raise money at $50 million. On the other hand, we have an acquisition offer at $45 million.”
VC: “Why would you even consider the acquisition offer?”
“Because I own enough stock that, after I pay my capital gains taxes, I will have enough money to do whatever I want for literally the rest of my life.”
“But I won’t be able to do that unless you sell for a lot more money in a few years. Think! I get 20% of what my fund makes, but you guys are just a small percentage of our overall assets. Plus, my personal burn rate is about 10x higher than yours. You should really keep plugging away.”
Most of the time, venture capitalists give good advice because they know what they’re talking about and they have the right incentive. But in this particular cases, the incentives are just slightly misaligned, and VCs have a systematic bias towards going for broke.
Basically, the VCs need to be more like Tom Pollack at Universal: take less risk upfront in exchange for giving up more in the future. A high liquidation preference does exactly that: it means VCs get a bigger piece of the immediate upside, at the cost of a smaller piece of the future upside. You’d see fewer cases where someone quits a bigco, starts a startup, and gets acqui-hired by another bigco (or even the same one). The startup is basically a way to request, and get, a big bonus. You’d see more startups willing to aim higher, since that’s the best way for the early employees’ work to be worth it. At the same time, VCs would be more willing to accept early acqui-hires, and less monomanic about pushing for unicorndom.
Would VCs accept that trade? In theory, they’re all hunting for the Next Big Thing, but 1) that’s because that’s historically how the money has been made, but if liquidation multiples went up, that would change, and 2) they kind of have to; VCs are middlemen between startups and limited partners, and in effect they work for whoever provides the scarcer resource. (A mini-history of the last ten years of the venture business: the people who were the first to realize they were working for founders, not LPs, made the most money.)
Why It Might Not Happen
The venture capital business is between 70 and 4,000 years old, depending on whether you start your counting with Arthur Rock or the ancient Mesopotamians. The rules have a short half-life. Right now one of the rules is that founders have the upper hand, so they call the shots; capital is everywhere, talent is rare.
But that’s getting less true. Distribution used to be artificially cheap, but now every scalable channel is controlled by one or two extremely savvy companies. These companies are mature, in the sense that their growth is driven by ratcheting up price and ad load more than from growing usage, and ad load is a tax on organic user acquisition while ad price is a tax on the other kind.
There was a wonderful time in the mid 2000s when starting a company was about as cheap as being unemployed. When you’re using open-source software and your founders live cheaply, a small early round can go a long way. And of course, cool stuff spreads by word-of-mouth online.
Have you noticed a significant drop in the volume of cool stuff spreading by word of mouth in the last five years? I have. It seemed to happen around when “word of mouth” started to mean filtered one-to-many messages on feeds, rather than one-to-one messages by email. Once it’s happening on the feed, word-of-mouth advertising is a direct substitute for paid ads, so if whoever lives off the paid ads also controls the organic distribution algorithm, there will be problems.
(The major web platforms all argue that there’s separation of church and state between advertising and organic content. I want to believe it. But even if there is, there’s a dog-that-didn’t-bark problem; advertisers get customer service; growth hackers get trapped in spam filters. You don’t need coordination of efforts when alignment of incentives is hard at work.)
Another problem: the industry is even more synonymous with the Bay, which has basically not built any new housing since it became the global headquarters for the world’s most important industry. As a rough approximation, standards of living for new founders are about what they would have been ten years ago. Their paychecks are much higher, and that’s all been absorbed by landlords and being in a higher tax bracket. California’s economic management has evolved to subsidize things being descended from people who owned farms back before Silicon Valley was Silicon Valley, and to pay for this by taxing their most competitive industry.
California still benefits on net from the agglomeration effects of having a huge proportion of the world’s smartest programmers all within Lyfting distance of each other. But those benefits more directly translate into revenue for bigger companies, which has put upward pressure on the entry-level programmer wages that now set the floor for the price of a two-bedroom non-rent controlled apartment in the bay area. For startups, agglomeration effects are a theoretical benefit to market value and a definite hit to overhead.
Fixed costs are higher, marginal costs are higher, and it’s harder to fly under the radar. There’s less margin for error,thus less room for creativity. Maybe, after a decade-long run, the pendulum is swinging back towards capital, at least a bit.
But just in case it’s not, we really should try to bring back high liquidation preferences, at least some of the time. Come on, Silicon Valley, are you really going to let Hollywood, of all places, be more sophisticated?