A study I'd like to see done | Hacker News Reader

A study I'd like to see done

7 pointsdfranke29 comments

Ask the following question to twentysomething males whose net worth is not currently above $250k:

Suppose someone offered you the following wager: consider your net worth on the day that you turn 55. If it is greater than $n, where n is adjusted for future inflation, then your worth doubles. Otherwise, you lose all but the last $5000 of everything you own. If you are in debt, your debt doubles. What is largest value of n for which you would accept this wager?

I'd be fascinated by the results of a controlled study which compared how answers to this question varied across cultures. Anyone want to posit any hypotheses, or give their personal responses to the question?

Edit: perhaps the question should be changed to "consider the highest that you net worth has ever been...", to accomodate the case of people who have already gotten very rich but given most of it away.

View on HN

For me: n <= ~10,000.

This isn't a very enticing wager--it seems like I have a lot more to lose than to gain. If I set n at, say, $1M, then I might win $1M, which would be nice, or I could end up with basically nothing, which would be devastating. You might have a more interesting wager if $5000 were changed to something more substantial, like $250k.

> If I set n at, say, $1M, then I might win $1M

I'm not sure whether you're understanding the wager correctly. If you win, you double your entire savings, not just gain $n.

Oh, that's true. Still, I think the wager is too scary for most people to accept for large values of n. Even if I set n=1M and have 10M when I'm 55, gaining the extra 10 million will make significantly less of a difference to my life than losing all but $5k will.

that's the point of the study

Dfranke, adhering to the constraints given shouldn't the correct answer should be a very large negative value for n? There's no upside to increasing the value of n; you're only putting yourself further at risk.

I'm asking the maximum value of n you could be offered such that you would still accept.

Yeah. Marginal utility has to be taken into account as a confounding factor.

It's only a confounding factor to the extent that the shape of the curve varies across cultures. The only reason that comes to mind for any such variation would be the availability of charitable handouts or government welfare, and that's objective enough to easily factor out.

There was a simpler question about risk on the first Y Combinator application: http://web.archive.org/web/20050324021230/www.archub.org/sfp...

>If you could trade a 100% chance of $1 million for a 10% chance of a larger amount, how large would it have to be? Answer for each founder. (There is no right answer.)

Quite a study. I'd like to see a study in which a group of very young children (even babies) are made to populate a controlled environment (where they are served by Automat-style food dispensers) in order to see what form of government they develop in such an untarnished state of nature.

My guess? Anarcho-syndicalism tempered by naptime.

My guess is something closer to Lord of the Flies.

Why would it be different than the societies we have?

I read something like that sometime this year. It was kids of age of the kids from Lord of Flies and they gave them an island or something similar, I forgot the results though.

My guess? Tribalism.

From what I understand, there is no way they would develop any kind of compex language in one generation, and would live something like apes do.

Go watch Kid Nation. Wednesdays at 8:00 on CBS.

21 Male/USA - Depending on how the wager is presented, I could probably be sold on $n = $1 billion.

For example, one should consider it a motivational tool - sure, most people in my position could make $1 billion with the right motivation, and this bet is just the right kind of motivation!

To take such a bet is morally required, because it'll lead me to leading a life where I create more wealth for humanity, even if I don't reach the $1 billion mark.

So morally, the question shouldn't be about maximizing one's outcome at age 55, but about finding a value that I would deem achievable for the next 35 years.

Also, even if I don't make $1 billion, the kind of skills I'd learn trying to make $1 billion over 35 years would allow me to live a comfortable life after 55 anyway.

Not high. Money has diminishing marginal utility. If your net worth is high, you get little extra benefit for doubling it.

So I'd say $100k, just off the top of my head. Seems like risking 95 to win 100 is pretty good, not to mention it's trivial to get there.

If it's trivial to get there, then why wouldn't you be willing to go higher? Remember, you double your entire savings if you win. No matter how quickly your marginal utility curve flattens, it seems absurdly risk-averse to name any number lower than half of whatever you think you need in order to retire.

I'd need about two mil to retire comfortably I'd say. It's far easier to get to that from one million than it is to get to one million from $5,000. Hence I'd be seriously screwed, especially at that age, If I found myself with $950k and had to lose it all, whereas had I not made the bet, I'd probably be at most 5 years from retirement.

Two million to retire comfortably? That's quite a standard of living. Supposing you're earning 6% interest and expect to live to 100, that's a budget of $10725 per month. That's enough to spend the rest of your life living in the penthouse suite of a cruise ship.

Right. Exactly

A retirement on half as much income as you'd like seems a lot better than a retirement on $5000...

Well, it has the problem that people may answer surveys differently than the same question in real life.

It looks like it tries to expose expected future earnings and risk (avoidance|tolerance). Did you mean that? If so, that relationship may show in asset allocation choices by people over their lifetime. Data from any tax firm would work....

Also, why focus on 20-year-olds with less than 250K? Have you already established that current net worth and/or age has a quantifiable impact on risk aversion?

Patrick

I'd say about $900,000 ( in 2007 dollars).

My logic is that I'd be approaching retirement age. I could/would definitely plan my career so that I had enough money to retire comfortably, which would require about $1.5 million. I'm saying less than that just to be safe (and in case I really want to switch to a less lucrative/more fulfilling career). In reality, I think I'll make a lot more than $1.5 million, but it's a little too risky to bet on.

Heh, so among a relatively homogenous group, the first two answers I've gotten vary by a factor of 90. Nice.

can't i beat the wager by getting my parents and friends to give me a big pile of money a few days before my 55th birthday?

Not if you owe it back to them. I said net worth, not gross.

He said give, which means no obligation other than to the federal government.

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

thread winner

Reply on news.ycombinator.com