Hacker News Reader

Comment by code4tee | Hacker News Reader

Easiest strategy is to just assume the options are worthless and base your comp assessment on that. Doing otherwise sets people up to get burned badly. Furthermore the risk reward for all but the founders is typically significantly lopsided.

If you can accept the risk great, but again assume you’ll never see a penny from options or far far less than you might think (as the calculator highlights nicely).

If things go well you’ll get a nice bonus. Not life changing for the vast majority of people, but a nice financial surprise.

The mistake most people make is accepting far too little cash comp on the grounds that their options may be worth something some day. When they turn out not to be they get burned twice. First on not getting that money period, and second on not having higher cash comp all along which means they also missed out on compound savings or investment with that money.

Net net see options for what they are in most companies—-a way to “pay” people when the company can’t really afford to pay people.

View on HN

That’s not a rational way to evaluate compensation. Would a reasonable person rather take a guaranteed $1000 or a 10% chance of a $1 million? Obviously the latter, even if it has a large chance of being worthless. The not-stupid way of assessing a compensation package would be to conservatively estimate the expected value of the options and apply a substantial risk penalty in your objective function, not to round all volatile compensation down to zero.

Or more realistic: Would a reasonable person accept $1000 a month or a 10% chance of a million dollars?

... without you knowing whether it's a 10% chance or a 0.2% chance and that number becoming 0 if you leave the company so you stay inside and spend even more time working for them even though you might get more at another place, or get a second shot at least. Big tech has multiple advantages, including higher total comp and flexibility.

That’s not a rational way to evaluate compensation. Would a reasonable person rather take a guaranteed $1000 or a 10% chance of a $1 million?

A "reasonable person", or a more commonly used "rational agent", is an idealized notion that doesn't actually exist in a real world. Actual humans are not expected-utility-maximizers. See e.g.[1].

Also importantly, the scenario (evaluating the value of option-based compensation) is not like "guaranteed $1000 or a 10% chance of a $1 million" scenario you're describing. Rather, it's like "pay $50-100k/year in opportunity cost for entry to the lottery in which you can win some unknown prize that's almost surely below $5M, at even more unknown probability". Even an expected-utility-maximizer, a "reasonable person", cannot really make a reasonable calculation of the expected utility.

[1] - https://sci-hub.tw/https://www.sciencedirect.com/science/art...

Just because people don't tend to do it doesn't mean you as an individual are incapable of doing it. His advise is the rational approach, and people on Hacker News tend to be smart enough to understand how to estimate probability distributions and how to compute expected value off of these estimates. With tools like this at their disposal, they are more than capable of applying their risk preferences and making rational choices.

It isn't really a matter of being "smart". Even investors, who have detailed financial models, decades of data for comparables, get preferences and often board representation to protect their interests, aren't confidently going to be able to provide a probable value of the investment. It is certainly unreasonable to expect employees, who get no board representation and thus can be screwed over in any scenario short of an IPO, to do better than people who have this as their entire job.

That's precisely right. Calculating EV of a lottery ticket is easy, because you know the odds exactly. The idea that a "smart person" can do the same for an early-stage startup that's offering options to employees simply beggars belief.

Estimating the value at zero is no different. I don't know if options from my old job will ever be liquid and worth anything of note, but I wouldn't sell them for a penny.

His advise is only smart if the salary you'd be giving up is unnecessary and you consider it a luxury to lock into an investment you may lose.

In fact, that's exactly what it is, whenever you're taking stock options into account for compensation be sure to consider how big of an investment (in loss of salary) it is and whether the potential payout and the risk seem logical. If you think that making that big of an investment is too risky for your well being (even if, in the average, it'd pay out better over a lot of plays) then don't take it.

We exist for a limited amount of time and large games of chance like this (even if they have an expected return) are not irrational to decline.

I cannot think of a single employed person that I know for whom turning down a $1000 1000:1 payoff for 10:1 odds is rational. Maybe a recovering gambling addict who would be triggered by this action, but that's about it.

Turning down an offer can be perfectly rational. I do not know why I am being accused of advocating for always taking options. All I'm claiming is that they aren't always worth zero, and assuming such is an irrational decision. And it's an entirely avoidable irrational decision that can be addressed by considering your own risk preferences, foregone salary, tax consequences, expected company outcomes, professional development, &c.

There is so much to factor in, too. I'm pretty happy with options if my base salary gives me a decent quality of life, I certainly wouldn't work for less on a gamble, considering the failure rate of startups. Never mind the fact that you are making a secondary investment in the same place you work, so you are doing worse than putting your eggs in one basket.

It's a nice bonus that _might_ pay something out. It's not a replacement for a proper salary.

That said, I can appreciate the desire to get skin in the game. Would love to see research on how necessary that is, though.

I think you may still be missing the point. What is the harm in completely ignoring the stock options and negotiating strictly on base pay. I think stock options for the most part should be considered like free snacks in the workplace. "Nice-to-have", they taste better because they're free, and if all the sudden they get yanked you didn't stake any livelihood or change of circumstance for it. If they end up becoming something great - maybe you get a new espresso maker, extra bathrooms at work, or an office instead of an open desk. Rarely will stock options change your life that you shouldn't plan on them but a nice perk if they do as they are very infrequently life-changing by any measure.

> His advise is only smart if the salary you'd be giving up is unnecessary and you consider it a luxury to lock into an investment you may lose.

Actually, even then, i'd just take the cash comp instead and invest it into FAANG options. You can do that on almost any brokerage.

> people on Hacker News tend to be smart enough to understand how to estimate probability distributions and how to compute expected value off of these estimate

That's a strange assumption to make. If anything, I'd argue that people who think they are somehow smarter than others because they're an engineer/programmer/whatever are more likely to put too much faith in their intelligence and make financial mistakes.

> His advise is the rational approach

Being smart is orthogonal to being rational.

What is the probability distribution that startup X will succeed? I'm genuinely asking, I'd love to know how would you approach estimating something like that?

Building this I used information from CB Insights, which looks at companies that receive VC funding by cohort. Their most recent report covers companies that raised seed rounds in 2008-2010: https://www.cbinsights.com/research/venture-capital-funnel-2...

>> more than capable of applying their risk preferences and making rational choices

Only if you know 1. The Cap Table 2. Your founders' and boards' predilections for exits 3. Potential future rounds 4. Where the market will go

When you don't know these, you can still create a probability distribution, but price will likely be zero because the inputs have to have massive brackets for potential values.

Another input I would add to your list is knowing what the day to day culture of the company is. If it's a total shit show and many many startups are, you might not even be able to last there a year which is the normal cliff before your options start vesting.

>"Only if you know 1 The Cap Table

Are candidates or regular employees ever privy to the cap table? I have asked about it many times when a recruiter has tried to sell me on the joys of options in lieu of pay. Even after accepting a job I have routinely been rebuffed on getting a response to inquiries about the cap table.

Early stage key hires are; after things get rolling though probably not so much.

Don't forget liquidation preferences.

This isn't even an abstract question about utility. What esoterica outlined was finance 101 (change "compensation" for ROI), which millions of people use in their day to day lives to make investment decisions. And stock options are an investment just like any other.

In my limited experience, stock options have significant differences from other typical investments in the following ways: 1) You can be prevented from effectively selling stock options prior to a liquidity event 2) Partly due to 1) , stock options are much harder to price than other investments that _could_ be bought and sold freely. 3) Due to the additional requirements as a byproduct of vesting, it is often impossible to pursue options at multiple similar organizations simultaneously. 4) Due to the caveats of even being a shareholder in a company, sometimes there are complications and risks. This is why there is a significant multi-page document to sign when exercising options typically.

What is a "typical" investment? There are literally thousands of different assets one can invest in, all of which contain varying degrees of risk profiles. What you outlined was the risk profile of options, which you would use to determine ROI.

I’m reasonable. I would never take a lottery ticket vs $1k in cash unless I could buy 20 tickets.

> Would a reasonable person rather take a guaranteed $1000 or a 10% chance of a $1 million?

For a lot of good software engineers, it's more like:

Option A: $1.5mm (5 years total comp at FANG)

Option B: $0.5mm (5 years startup salary) + 0-10% chance of $0-20mm.

Given all the unknowns in B, I'd definitely take A.

I'm on the Option B track right now :) And there are a lot of unknowns. But I would say that I'm "learning" more at B then I would at A.

I can't speak for Facebook, Google, but I did work at Amazon for about 5 years (and take this with a grain of salt because this was 2003 through 2008 on the ops side (supply chain specifically)). And you learn the Amazon way of building software. You use Amazon tools, frameworks, and style. And while some of that knowledge is definitely applicable to other jobs I've held, most of it is not. It is a little bit more now because of AWS (similarity between internal tools and AWS offerings).

So yeah...with Option B you might not hit the lottery. But the skills I learned from multiple attempts at Option B are more applicable in my opinion. And that's worth something.

Yeah. Either one is a good salary. At some point enough is enough. Take the job that you will enjoy. Even as I say that, I know that for some people job == money, so if making the most money is "fun" then that's totally OK.

"I took the one [road] less travelled by, and that has made all the difference" ;-) It doesn't matter which way you go, years from now you'll end up thinking the decision was an important one.

I disagree. Either one is a good salary if and only if you are willing and able to work for the rest of your life. I would spend 15 years at a FAANG and retire early with a comfortable amount of money.

>So yeah...with Option B you might not hit the lottery. But the skills I learned from multiple attempts at Option B are more applicable in my opinion. And that's worth something.

Obviously speaking from my own experience, but this hits the nail on the head. The things I learned at startups were parlayed into jobs at the bigger companies later in life.

Also, on the original topic. I've been a part of three startups. One with trivial equity, one with options worth about 0.1% of the company, and finally one with options worth 1% of the company.

The second company is still going and is a lifestyle business, so those options were basically worthless since the owner may not sell for many, many years. The first and third companies were acquired, but at values that made the options worthless, so I got a long term capital loss for the first company and nothing for the third.

For the third, I was an executive, so I got a stock/cash package from the acquiring company for (very) low seven figures, but the stock itself was worth nothing.

So, I'm 3 for 3 on "successful" startups, but 0 for 3 on actually cashing in on any equity.

I would be really curious to get an answer on these, for the company you had the executive role:

- Was it basically an acqui-hire?

- How much was it sold for?

- How much was raised totally?

- How much was the cumulative compensation package above the stock price given to key employees such as yourself?

Thanks

If we were talking about horse racing, I’d agree with you, as you can evaluate the odds of an outcome.

The reality is that in business as an employee you’re ability to project the odds is limited at best. You simply don’t have the information to make that calculation.

Obviously there’s a spectrum of risk, which is why most people don’t work for minimum wage plus options. But it’s pretty clear that in a world that hasn’t seen a tech recession in a long time, employees without the perspective of living through 2000 may not appreciate what can happen, and will happen again.

If a rational person was given this choice 100 times it would be rational to always chose 10% chance of $1 million, but is having 10k cash in your pocket 10 times as valuable as having 1k in your pocket when compared to having 0 in your pocket?

Money has diminishing returns and when you're potentially going to be below water when it comes to mortgage, food, utility, even the occasional vacation for sanity - then a smaller guaranteed sum can be more valuable than a larger expected sum.

That depends on the marginal value of a dollar. If you don't have any money, a rational person would take the first $1,000 since "some food" is better than "a 10% chance of some food".

Your example dramatically overstates the upside, understates the risk, and ignores a lot of the difficult problems in evaluating options - there are large uncertainties in valuation and very limited information, there is a negligible possibility of fraud or misrepresentation, the person evaluating might have good reason to choose less risk (e.g. starting from 0 savings greatly increases the value of stable predictable income), there is additional risk in having salary and investment tied to the same company, etc.

Let's change it to make it slightly more realistic - on offer are options which have a 1% (at best) chance of being worth 1 million in 10 years or a guaranteed higher salary over 10 years worth 300k more. It's not quite so clear cut in that case, and depends greatly on the circumstances and judgement of the person making that call with very limited information.

I think it's more reasonable to assume they are worth 0, as the majority of options end up at 0, and the probability of making more through options than salary is nowhere near that in your example.

I don't think your example depends on the person's circunstances, there is a very clear cut rational decision there. Even not taking the utility of money into account, nor the fact that you need to discount that million as it only arrives in 10 years, you are still only making 10k on average on that gamble. Taking the 300k is the rational decision.

A reasonable person would take the lottery ticket and immediately offer to sell it back to the issuer--or to a large insurer or casino--for a guaranteed $50k (negotiable) before the drawing occurs.

However, some options grants are not transferable in that fashion.

The reasonable person can then calculate the range of payouts for each option as [$1k..$1k], ev $1k, for the cash option, and [$0..$1M], ev $100k, for the gambling option. Not all gamblers only use expected value as their only metric. Some people also use the minimum return. Those people would take the guaranteed $1k, in cash, and walk.

Even if the gambling option had a minimum payout of $1000, thus making it the strictly superior option on paper, just by those metrics, it might also only pay off 3 days from now, when the $1000 on the spot, could be used right now, possibly in some other psychologist's thought experiment on gambling behaviors.

Because it is more like $200K/year and 0.1% chance of $500K or $75K/year and a 0.1% chance of $5M. It would be silly to take the latter over an expected 5 year tenure.

> Would a reasonable person rather take a guaranteed $1000 or a 10% chance of a $1 million?

When they have monthly expenses, or understand the power of compounding interest? A reasonable person would take the guaranteed.

----

Even if you make your example less hyperbolic (for an employee):

$100k vs $90k + a 5% chance of making $50k in 5 years.

Where, the higher and guaranteed salary also has 3% annual COLA (or further freedom to job switch) raises, has a 4% 401k match; The wise person would take the guaranteed income.

Expectation of option B is $100,000 - so any rational human would take that.

The numbers I've seen are more like: give up anywhere from $30k to $130k per year for a chance at an equity payout. The only way that works out is if $num_years_worked * $discount_salary < $potential_payout * $chance_of_payout. So if you are getting shorted $50k/yr and you spend 8 years helping build the company, you need a payout that is better than $400k with 100% chance of success. Something like 90% of companies utterly fail. I'm sure some VC math can be applied here. But the VC gets to roll the dice a lot. You get to role the dice once every few years as you change jobs after getting some vesting going on. There is a reason why it is called the start up lotto. You can pad the chances by going with founders with a track record, or going for a company that is nearing an exit, but it is all still the start up lotto.

I think the difference is that equity has the potential (even if it’s a small chance) to create a major qualitative change in your life. If you stand to make a million over 8 years anyway, another 400k is nice but won’t change your life that much. Maybe you buy a slightly bigger house or you retire a few years earlier. But a homerun outcome in a startup can mean complete financial freedom and never needing to work again. It’s the long tail, sure, but given that there are other benefits to a startup (learning more, more influence, etc.), a lot of people prefer that long shot to the certainty of being marginally wealthier.

I also don’t think it’s really comparable to the lottery. It’s like playing the lottery if you could easily afford to have a 1-5% chance of winning... in which case it wouldn’t be such a bad idea to play for many people.

Different strokes for different folks and all that, but I do think it’s wrong to imply that it’s always wrong to take a calculated risk like this when given the opportunity.

> I also don’t think it’s really comparable to the lottery. It’s like playing the lottery if you could easily afford to have a 1-5% chance of winning... in which case it wouldn’t be such a bad idea to play for many people.

Really? Roulette has odds of 37:1 for a single number which is about a 2.7% chance. Would you take $300k that you could otherwise earn in salary and place it all on a single number? The payout would be 10.5 million which far more than anyone I know expects to get from their stock options even with a pretty good outcome. You'd still be insane to walk into a casino in Vegas and make that bet.

I would, assuming that if I lost:

- I could pay off the 300k over the next 8 years.

- The loss would be tax-deductible.

- I could somehow know for sure I'd have a job where I'd still net at least 120k per year during that period (after the loss).

- It would somehow grant me multiple years-worth of extremely valuable skills and experience, new friends and connections, and the chance to be intellectually challenged and satisfied.

I think that would definitely be the most popular roulette wheel in Vegas!

> Would a reasonable person rather take a guaranteed $1000 or a 10% chance of a $1 million?

Can this reasonable person afford to pay this month's bills?

Idk, does taking the second option mean I have a 90% chance of not making rent?

I think you are misreading the point. It's not that the true expected value of early stage options is actually zero, it's that it might as well be for your purposes, choosing between that and other reasonable (i.e. market value) opportunities.

If you look at early stage startups, I think it's pretty clear that the EV[startup options] < EV[invested market value salary delta]. This holds true over nearly all startups, and nearly all people who have the option of early stage options. There are outliers, sure - but to a 1st approximation you aren't in them.

So you have to fall back on "how much more fun/cool is this startup job than other things I can do". And realistically, getting rich off a startup isn't a rational plan for nearly all people. Asking yourself, "what are the odds I recoup most of the lost salary?" is a more like it. And for typical seed round start up, that break even is going to be a few percent, dilute as needed to look at different rounds.

Of course this doesn't help you choose between two different under market salaries with options, either...

For that to be "obvious", the 10% estimate should be objective and not just pulled out of thin air by somebody. And if it is statistics-based, it should hold into the future even though conditions change all the time (no idea how you would guarantee that). Even then, it's not that obvious - you can't pay rent with a 10% chance. Surely, if you take such chances regularly, it would pay you rent on average, but if you do it just once, you'd have to think about how much not being evicted from your apartment when 90% chance plays out instead costs to you.

The problem with that logic is that you never know up front what the percentage chance is, and whatever it might be, a retrospective analysis shows it's always much, much lower than 10%.

A conservative estimate of the expected value of the options in a startup that is not successful yet, is $0.

So assuming the options are worthless is not all that unreasonable. Options are bonus, not part of the real compensation. Unless the company is already successful or your share in the company is large enough that it really is worth the risk.

But if in the most optimistic scenario your options end up being worth $1 million, but it's going to be years before they're worth that much, the chance they're going to end up worth that much is small, and in the mean time you're severely underpaid, then it's probably not such a good idea.

I would also add that stock options are not stock grants. You are given the option to purchase those shares. Meaning you have to pay the company for them at strike price within 90 days of leaving and you have to pay tax on them.

Many young startup employees I talk to are under the impression that they own .x% of the company because their options have vested. That is not true.

After exercising your options the company could shut down or it can be acquired with a valuation less than what your options and strike price were based on. In both of these cases you would be losing money.

Grants would need to be taxed at fair market value even though it's not liquid, so there's a very good reason for doing it this way

The cost shouldn't be 0% or 100%, but the percentage chance of getting the reward, multiplied by the value of the reward. IOW, E(X) = P(X) * X, so a 50% of 100K is effectively a 50K payout. The problem is P(X) is unknown, and you only really have heuristics to guesstimate. Whether you move forward with the idea 0%/100% is just if you are a glass half empty/full kind of person, but the real answer is somewhere in between.

From looking at tldroptions.io, it's just taking an empirical approach to guess the P(X) value.

> Easiest strategy is to just assume the options are worthless and base your comp assessment on that.

But that easy thing leads to:

a. Having no basis to negotiate your options. (X% of $0 is always $0 regardless of X)

b. Your comp will always look like crap compared to a publicly traded company.

Then negotiate a higher salary which is the statistically correct course of action. Many of my colleagues are 10+ years stuck waiting for an IPO that may never come.

B is kinda true a priori. It's easier to negotiate A should be negotiated in relationship to ownership and responsibility. You should get into the head the hiring manager in question and figure out what's the max/min range to motivate a candidate to do the job properly. That could be anything for single digit percentages for senior ICs at an early stage company to 50% for a pre funding cofounder to a couple bips for an IC at a series D medium sized company to a sizable fraction of a percent for an executive at a series B. Trying to value illiquid options as if they were fungible in cash is a fool's bargain. Reject it. They have value, but they're different.

In your example, what would you consider an early stage startup? At what size org would 20 basis points be reasonable for a senior IC?

Sorry, I should have been more specific -- I meant seed stage. I think you'll see 20 basis points for a senior IC around post Series A/B depending on the company and how they've set up their option pool, as well as the senior IC. At that point, 20 bips is material enough to be non-negligible if the company grows well enough to the next raise, but probably not enough to be thought of as cash in the lifetime of that employees tenure (on average, anyhow). At that point, cash comp should be on par with median salary, which is tier 4 and below for most ICs*

*I'm putting FANG into T1, the next tier down of liquid companies into T2, large pre-IPO companies into T3, and everyone else into T4 -- the interesting part obviously is that company salary tiers can and do obviously change if/as companies grow successfully.

Thank you!

a) That's not true. More is more. You just do it from a basis point perspective.

b) If you're at a pre-ipo company with options, your comp is crap compared to a publicly traded company. If you're counting your lottery tickets as real money, you're a fool. You can't even sell them on Sharespost for the valuation the company tells you.

Yes, on Point B the comp IS “crap” and hence why most startups try to convince you to take the options. They can’t pay you what the competition would give you in liquid comp so they try to get you to take the bet. Such startups are typically paying well below market on liquid comp. it’s not like they’re asking you to make a “bet” with a 10% haircut on liquid comp (which would make more sense).

I’m not against options, but if what you will make over the next X years based on liquid comp is not something you are totally OK walking away with as your total comp over that period then one is probably making a bad choice.

But your comp is almost always crap compared to a publicly traded company. (i.e. FAANG) Just look at the numbers on this website to get the expected value of startup stock. Short of getting an obscene amount of stock in a startup, it's not worth it. And almost no startup is going to give you 5-10x+ the amount of stock that other employees are getting for the same position.

Many people aren't comparing to FAANG companies: they are comparing to other places they actually want to work.

FAANG companies aren't paying that much because they get so much more value out of employees: they are paying that much more because that's how much extra they have to pay to convince people to take the jobs they are offering.

Obviously if they paid less they would get fewer engineers, but there are plenty who would take peanuts to work on problems at that scale. The actual reason they're paying that much because they have no limit on the number of 1% engineers they can put to work, and they want as many as they can get.

I think the subtext is that you're always taking a risk with start-up options, so minimize the risk by assuming the worst. Your basis for comparison/negotiation should be other offers, or your existing job and other offers. In a vacuum you'll never be able to make a realistic evaluation.

That is really great advice unless the company you're working at turns out to be really valuable, in which case it's really bad advice.

Only if you got in really early and even then it statistically won’t happen.

It’s like saying lottery tickets are a bad investment idea. It’s true for far more people than it isn’t.

It's <50%, but chances of winning the lottery are 1 in 292 million.

I'd bet if you picked a promising company (say Series B+ on the breakout list) your likelihood of a positive outcomes is >=10%.

Not high, but very different than lottery tickets.

You'd be betting on the probability of outcomes of choosing startups with successful outcomes, you must see the absurdity!

Regardless, at least a lottery's odds are fixed. For a startup not only do you have a direct impact on the outcome, but there's near-infinite number of internal and external factors that can make a company fail, or, very rarely, succeed.

I think that's one of the big selling points on options: It's a lottery ticket but you are directly capable of affecting your odds. Theoretically if a group of 100 people all feel that way, they should be able to do some amazing things!

Also, you can buy as many lottery tickets as you can afford - you only have so many shots at picking a promising company in your working days.

Those seem equivalent, just with a different limiting resource.

Even at seed stage, if you're picking from the same group as Crunchbase is the 2008-2010 cohort had a 30% chance of an exit and a 1% chance of becoming a unicorn. 1 in 100 is much better odds than 1 in 292 million.

So no one should ever work at a startup?

From a financial perspective, no. If you get more meaning from it and/or want more responsibilities and autonomy (on average), sure,

welcome to the HN paradox.

It's not that you shouldn't work at a startup, it's that you shouldn't let the startup convince you that their stock is going to inevitably be worth millions so you should accept a tiny salary up front.

Don't be Jack from Jack and the Beanstalk -- don't trade your cow for magic beans if you need to eat. Only make the trade if you won't be worse off if those beans don't turn out to be magic.

Reply on news.ycombinator.com