Everyone's monetary needs and goals differ. As professional poker players, there are significant advantages to several landmark bankroll numbers, such as(roughly):

20k for 2/5 nl
60k for 5/10 nl
150k for 10/20 nl
400k for 25/50 nl
etc

The expected utility of money has a somewhat sharp decline under these price points because of the ability to make more money by playing in bigger games when they are worth playing. You should keep reasonably safe investments with enough to play in order to keep your future earning power high. With additional money, you can take drastically higher risk investments, given that they have larger EV payoffs.

Some examples(ignoring inflation) in order of increasing risk are:

Cash (0% EV)
Bonds (4.5% EV)
Mutual funds (8% EV)
Real estate (~15% EV, not sure what number to put here)
Individual stocks (8% EV)
Cryptocurrency (50%+ EV, very debatably much higher)

As poker players, we need decently large reserves of cash to play regardless, so that takes care of a large portion of our needs. Additional money to put us over our bankroll requirements should likely be kept in mutual funds or real estate. Any additional money should be kept in the form of a well diversified cryptocurrency portfolio, given that the individual also believes cryptocurrencies EV far exceed that of traditional investments.



A graph of utility to money would look something like this, with the slope of the utility line decreasing slightly past every "landmark" bankroll point.

Am I thinking about this correctly? Anything to add?

Poker bankrolls should operate much more fluidly than that. Poker bankrolls should only be a static amount once the poker player has reached his skill cap, or is playing as big as he can play. Otherwise, there should frequently be shot-taking to higher limits in juicy games as long as the player can emotionally handle himself (and if he can't, he should get a new job immediately).

Your chart says that $120k has less utility than $0, which is obviously wrong. What you're actually talking about is marginal utility, ie. the slope of a utility function, which would gradually decrease as you accumulate more money. Your utility function should always increase with more money, but approach a flat line as you become increasingly wealthy.

Aside from that, the EV figures you've used are of course debatable (I would call cryptocurrency a zero EV proposition at best) but that's tangential and not really relevant to what you're asking.

Suggestion:



Juk

OP, you're not making any sense. And some of your return assumptions is questionable.

I think what you're looking for is Modern Portfolio Theory. The shortfall of MPT is that you need covariances for each pair of assets. If you get these empirically, but the problem is correlations are hardly ever static. Another approach is assuming correlations.

So you're saying if I invested $50K today and HODL until I'm 65 (32 years from now) you think my "investment" would be worth $21.5 billion??? And that's only at the measly rate of 50%. At 60% it'd be $170 billion. At 70% it'd be $1.1 trillion!

If I HODL a little longer (say till I'm 70, 37 years from now) 50% returns for $50K adding NOTHING would be $163 billion. 60% would be $1.7 trillion, and 70% would be $16.8 trillion.

I'm not sure your EV is totally accurate lol

50% wasn't debatable

+1, OP's graph is a plot of the first derivative of utility.

Going a bit deeper (perhaps unnecessarily), the plot implies that the second derivative is constant (or close to constant), but even that feature isn't necessarily true. For instance if most of your money is spent on charity - at certain (perhaps large) thresholds the first derivative (marginal utility) can be increasing.

Also, lol 50%

While 50% is very aggressive, to be fair, OP didn't say 50% for eternity.

Over the next year, there's a substantial chance of crypto tanking and losing the bulk of it's value, perhaps even to practically zero. There's also a substantial chance of it continuing it's massive upswing and conservatively 3x'ing in price. Even if those results are each 50% likely, that would be 50% EV. As someone else said, no I don't mean in perpetuity. This is largely an aside, and I don't want crypto to be the basis of this.

Thank you grim, I've been researching asset allocation trying to find more information on it. The modern portfolio theory articles on investopedia look like a decent place to start.

Any recommended reading on the topic?

The conflation of ROI (yearly) and EV here is amusing. The first post is equally amusing.