F*&k You Money
Why do you work? 1,000 years ago, the answer was obvious—do so or your family will starve in the streets. The industrial revolution left some adults forced to engage in the same existential calculation, struggling to keep the heating bills paid and the children fed. But for some of us, we are fortunate enough to have the opportunity to ponder how to make a life, buoyed by the assumption that we already have the capacity required to make a living.
And then, there are those who have attained an even loftier monetary status. For these individuals, choice of vocation is wholly severed from the compensation that employment confers. We romanticize and idealize these individuals. The millennials have coined the term “FIRE.”1 Their boomer parents would probably have chosen a more profane moniker (“f&^k you money”). But the implication is the same…earn enough to flip your boss the bird, drop the relevant f-bomb, and storm out of the office.
Is that the purpose of wealth? To remove the need for work? Or is the purpose to offer a greater array of choices in terms of what to work on or who to work for?
Developer Compensation
Developers, data scientists, and other technical professionals are among the modern age’s greatest beneficiaries. Their skills are in demand and the salaries they receive afford them both material comfort and the opportunity to invest in pursuit of a financially-independent future and an early-retirement.
An accomplished developer typically chooses among two primary compensation models. We’ll call them the “FAANG model” and the “startup model.” The former is fairly simple—offer the employee an extremely high salary relative to other competitors in the industry. The latter involves a decidedly lower salary, accompanied by equity. In the vast majority of cases, that equity is worth $0 to the employee, but within the distribution of outcomes is the possibility (albeit a small probability) of a life-changing exit.
Somewhere, there is a developer who was employee 19 at Facebook or Google or Amazon. That individual has amassed a net worth that no current hire of those companies will ever approach, regardless of the impressive salaries. Thus, the conventional framing of the choice between maximizing certainty or upside.
AE has created its own developer equity model. If you didn’t click that link and read the fascinating combination of economics, dad jokes, and marshmallows, the TL/DR is that AE offers its employees equity in projects incubated internally and from client projects where AE retains an equity interest. And even under conservative assumptions, modeling the value of your equity suggests that the projects AE incubates will yield considerably greater payouts than almost any alternative startup employment.
On the one hand, receiving equity in numerous projects rather than one decreases the maximum outcome (if one of those companies turns out to be the next FAANG company, your stake is smaller than if you were one of its employees). On the other hand, the probability of some of those projects yielding an exit is much higher than the probability of any individual project reaching an exit. Something about eggs and baskets comes to mind.
Simulation
One illustrative exercise is to generate some assumptions about the distribution of outcomes at startups, FAANGs, and AE, then run the clock forward hundreds of thousands of times.
So let’s make some assumptions about equity markets and our economy:
S&P 500 annual returns, mean ~ 8%/yr, stdev ~ 8%/yr
Perpetuity rate ~ 4%2
Inflation rate ~ 3%3
We’ll assume the standard tax brackets within the United States will apply to ordinary income along with 5% for state tax4, a capital gains rate of 20%, and a standard deduction of $12,500.
Now some parameters about income:
Financially-Independent Income requirement, per month ~ $7,500/mo.
Startup salary, annually ~$100,000/yr
AE salary, annually ~$100,000/yr
FAANG salary, annually ~$300,000/yr
Annual raise, annually ~3%
Annual investment rate, ~15%5
Initial net-worth ~$300,0006
Now comes the hard part—assumptions about the probability distribution associated with startup exits. Firstly, we’ll leverage some of the wisdom from techcrunch.7 Secondly, we’ll apply some fat-tail logic to the remaining 16%.
P(no exit) ~ 84%
P($50M) ~ 14%
P($500M) ~ 1.9%
P($5B) ~ 0.09%
P($50B) ~ 0.01%
We think our internal projects have much better odds, since we invest in Skunkworks projects where the first customer has been pre-sold, market fit is established, the founders already know they work well together, and a working MVP is launched8, but we’ll be conservative.
We’ll make some assumptions about equity granted to employees, and years of vesting:
At startups: 0.5% equity & 4 years of vesting
At AE, for our Skunkworks projects: ~20% of its equity is distributed to AE non-skunkworks-founder employees (the founder gets their own chunk!)
At AE, for some client projects: ~3% of its equity is distributed to AE’s employees
We’ll assume 3 rounds of dilution, at 25% per round.
Now we need to make some assumptions about AE, its growth, and how we will be allocating those employees:
Employee Growth, annually ~ 30%9
Skunkworks (with pre-sold customers) per employee, per year ~ 0.01
Skunkworks efficiency gain ~ 5%10
Equity-bearing client projects, per employee, per year ~ 0.1
So, starting with our team of 150, let’s run the clock forward 10 years, with 1,000 simulations… and explore the outcomes!
Results
Figure 1 - 1,000 trials, 10 years, and the quest for financial independence
Figure one illustrates the distributions of outcomes at traditional startups, FAANG companies, and AE. The black, dotted-line is the threshold for financial independence, given the desire for $7,500/month, in-perpetuity, after capital gains taxes, at a 4% withdrawal rate. Since a bunch of AE’s outcomes involve 10M+ (~4,000 trials or 4% of simulations), we have truncated the image for easier viewing.
Alternatively, for the mathematically-inclined, let’s look at the separately-generated log-scaled version, with 100,000 trials (with slightly higher parameters for illustrative purposes) in Figure 2. We’ll notice that the traces of results (the horizontal density bands near the top of the image) show some rather extraordinary results for those at traditional startups. Among those 1,000 draws are a handful of scenarios where the employee pockets nine-figure wealth. The AE simulations do not contain any such outcomes.
However, the AE simulation contains an enormous number of high-seven-figure net-worth outcomes and a significant handful of eight-figure net-worth outcomes after only a decade of work. Unsurprisingly, the FAANG outcomes are by far the most densely-packed. The variability is derived from market volatility. Bottom-line, if one earns that salary for a decade, a couple million in net worth is likely, but financial independence (especially if a few golden handcuffs are acquired along the way) is highly improbable.
Figure 2 - 100,000 trials, 10 years, higher-values log-scaled…
The expected log-utility of net-worth11 (Figure 3) is clearly leaning in AE’s direction, with FAANG outpacing startups by a smaller margin.
Settling a high bar and a short-horizon, not achieving financial independence is more likely than not. But at AE, your probability of reaching that end is an order of magnitude more likely than the alternatives. Sure, you might believe that you’ll have the discipline to work for a FAANG company, and live like you work for a startup, which would up your odds dramatically.
Statistically speaking, you will not.
Figure 3 - 100,000 trials, 10 years, log-scaled utility
Figure 4 - 100,000 trials, 10 years, will you be financially-independent?
We believe our best employees will reach financial independence, but rather than utter their “f*&k you” to us, they’ll flip their birds to the world of traditional employment and continue building amazing, agency-increasing products. But if your only aspiration is the ability to utter those syllables and ride off into the sunset, your odds are still notably higher here.12
Cynicism
Perhaps your initial reaction is cynical. In the parlance of Bayesian statistics, we would state that you possess a strong prior. After all, you’ve probably heard 100 promises from 100 startups, and perhaps only a couple of them ultimately came to fruition. In turn, you’re likely to disbelieve these values as well. Fair enough.
But remember, this model is not asserting that AE’s developers and data scientists are inherently superior (though some of the most talented of each are working here and we certainly think they’re as talented as any group working anywhere in the world). Nor are we asserting that our products are simply the greatest killer apps since DOS (though we are producing extraordinary products for ourselves and our clients). We are arguing that the incubation model we have created unlocks economic value that otherwise would never have emerged and that our compensation allows our employees to realize impressive gains in most of the simulated scenarios.
Oh, and if you find our parameters unduly optimistic, try your own. Explore our financial independence simulator!
1 “Financially Independent, Retire Early.” This group typically attempts to minimize expenditures and save/invest a sufficient quantity to allow relinquishing their day jobs.
2 In other words, what proportion of your assets can you spend each year without ultimately drawing down your account?
3 Everyone take a deep breath—this will not persist indefinitely. And if it does, there are larger concerns than financial planning
4 For those of you living in Texas or Florida, enjoy your absence of state income tax…. For those of you in California or New York…my humble apologies.
5 You would imagine the FAANG folks save a higher proportion, but lifestyle inflation is real. Might I suggest purchasing your car and your home at an early stage of your career… then increasing your salary… and having the audacity to retain the original pair of items?
6 Let’s assume, for the sake of argument, developers have a three years of salary in terms of net worth as they toil at startups. We’ll give folks the benefit of the doubt that they’ve invested early and often in their careers.
7 The article notes that the total proportion of startups that are ultimately acquired tops out around 16% (meaning 84% never exit).
8 How many startups can answer “yes” to all of those questions?
9 This is quite conservative, since AE employed roughly 25 folks a couple years ago, and has recently crossed 150 in terms of headcount.
10 As consulting revenue grows and our skills at incubation increase, the ratio of viable internal projects to employees will increase (as it already has). We’ll cap this at 0.05 for safety, but nonetheless.
11 Using a standard, linear expected value for net worth is common. It is also asinine. Are you indifferent between a net worth of $5M and a coinflip that renders your net worth $0M or $10M with equal probability? Of course not. Virtually anyone who is not a compulsive gambler would rather the certainty of $5M net worth. So then why, with large sums of wealth, do statistics professors insist on assumptions of linearity? Perhaps because they aren’t worth millions?
12 Though admittedly, if that’s your sole motivation, we’d rather you continue working at one of those places where everyone else also hates their job and logs on each day grudgingly in pursuit of their own visions of opulence.