NationStates Jolt Archive

I was writing an essay on behavioural finance the other week - which, if you're interested, is the use of psychology to try and predict how investors behave and then use the behaviour to make the most of investments - and I came across an economic theory which described a test to judge economic rationality. Economic rationality of investors is one of the assumptions (to a certain degree) that standard finance is based on, and behavioural finance challenges this by claiming that investors are irrational.

In this test there are two scenarios. Scenario A:

You have a choice of two options. In option 1, you recieve £8000 guaranteed. In option 2, you enter a lottery, with 85% chance of winning £10,000 and 15% chance of winning nothing.

Scenraio B:

You have a choice of two options. In option 1, you lose £8000 guaranteed. In option 2, you enter a lottery, with 15% chance of losing nothing and 85% chance of losing £10,000.

So which options from each scenario would you choose? Try to make your decision without looking up the economically 'rational' answer. If anyone's familiar with the theory, please don't spoil it until we hear what people have to say.

Scenario A: Option 1.

Scenario B: Option 2.

I never gamble to win money, only to lose it.

Though I suppose that is the reality of it all...

I know the answer, that's what you get when playing poker :p

So I'll just shut up and vote instead :D

A : 1

B : 2

In the first scenario, a possible $8000 plus a possible extra $2000 dollar profit is nothing compared to a guaranteed $8000 dollar profit.

The second scenario, if I'm going to be in debt, I don't see why I shouldn't chance my luck at clearing all of it for the potential price of having to pay a mere 25% extra.

I am assuming the economically sound answer is _____, but I won't say it.

In the first one, I'd choose option A. It's not worth risking 8000 to win 2000 additional pounds. In the second scenario, I'd go with the 8000 loss; again, the risk is just too high to justify taking the chance of losing 10000 instead of losing 8000.

I need to ask for clarification, but in doing so I might give away the answer.

Thus, the only winning move is not to play.

**goes to find a chess set**

I need to ask for clarification, but in doing so I might give away the answer.

Thus, the only winning more is not to play.

**goes to find a chess set**

Actually, you win more just by playing scenario A and then bail out :D

Actually, you win more just by playing scenario A and then bail out :D

True, but I don't know that this option is available. Thus, the next best answer is to stop wasting my time with hypothetical money and go play chess instead.

Depends how much money I have. The more money I have the less I have to worry about financial constraints and will then worry more about statistical outcomes.

Only way to truly know you'll come out on top is to physically beat the odds themselves. Thus, unless Chuck Norris will accept beer kegs from you since you're a close buddy, don't play.

I'd also like to add that the two scenarios should be considered totally independent of each other, rather than scenario B following on immediatly from scenario A.

I'll tell you the theory and its implications soon, after some more people have answered.

:eek:

I just remembered the "spoiler" tag. I'll use that to ask for clarification.

Cosmopoles,

Do the two scenarios occur simultaneously in time, in chronological fashion (Scenario A happens, then Scenario B), or are they completely unrelated and independent of each other in time and space?

The second option, as in parallel universes with no relation.

I'd also like to add that the two scenarios should be considered totally independent of each other, rather than scenario B following on immediatly from scenario A.


Ah.

Independent as in simultaneous, or independent as in each occurs in its own parallel universe with no relations of any kind whatever?

In the first one, I'd choose option A. It's not worth risking 8000 to win 2000 additional pounds.


In Scenario A, the worst possible outcome is 0. No gain, no loss. The best possible outcome is 10,000. The 8,000 is a red herring because it is irrational to throw away an 85% chance in favor of the best possible outcome for a sub-optimal outcome.

In the long run, you'll make much more money taking the 85% risk then you will playing it safe (assuming the Scenario occurs repeatedly, anyway...)


In the second scenario, I'd go with the 8000 loss; again, the risk is just too high to justify taking the chance of losing 10000 instead of losing 8000.

In Scenario B, the worst possible outcome is -10,000. The best possible outcome is 0 (no loss). The -8000 (first option) is again a red herring. Yes, there is only a 15% chance of getting away with no loss. But by choosing the -8000 loss outright, you are reducing that 15% long shot to 0% yourself.

Choosing to reduce a long shot possibility to a guaranteed loss eliminates outright the only means by which you could get away with no loss. Reducing your options in such a manner as to produce a guaranteed less than optimal outcome seems to be highly irrational.

The second option, as in parallel universes with no relation.

Righto.

**votes in poll**

I just read about something like this being tested on monkeys and they gave the same response we do.

Presumably they offered bananas rather than financial reward?

In Scenario A, the worst possible outcome is 0. No gain, no loss. The best possible outcome is 10,000. The 8,000 is a red herring because it is irrational to throw away an 85% chance in favor of the best possible outcome for a sub-optimal outcome.

In the long run, you'll make much more money taking the 85% risk then you will playing it safe (assuming the Scenario occurs repeatedly, anyway...)This is true. On an iterated experiment, you average wins of 8,500 per play rather than 8,000 in wins per play. 8,500 is better than 8,000.

In Scenario B, the worst possible outcome is -10,000. The best possible outcome is 0 (no loss). The -8000 (first option) is again a red herring. Yes, there is only a 15% chance of getting away with no loss. But by choosing the -8000 loss outright, you are reducing that 15% long shot to 0% yourself.

Choosing to reduce a long shot possibility to a guaranteed loss eliminates outright the only means by which you could get away with no loss. Reducing your options in such a manner as to produce a guaranteed less than optimal outcome seems to be highly irrational.This is wrong. On iterated plays you average losses of 8,500 per play, whereas taking the certain loss you average losses of 8,000. 8,000 is better than 8,500.

Clarification for Loyal Opposition's assumption:

The scenarios are not continuous, i.e. they do not repeat - they're a one off

The 8,000 is a red herring because it is irrational to throw away an 85% chance in favor of the best possible outcome for a sub-optimal outcome.

It's not "irrational" in any objective sense--it depends how much I value that additional two thousand dollars compared to how much I fear the risk of losing it all.

If I'm content with 8,000, if I'm not that concerned about the extra, there's no reason to take the risk.

In the long run, you'll make much more money taking the 85% risk then you will playing it safe (assuming the Scenario occurs repeatedly, anyway...)

Yes--if the scenario occurs repeatedly. We have no reason to make any such assumption.

Choosing to reduce a long shot possibility to a guaranteed loss eliminates outright the only means by which you could get away with no loss. Reducing your options in such a manner as to produce a guaranteed less than optimal outcome seems to be highly irrational.

Not at all. Just as with the first case, if I am content with losing the eight thousand but fear that losing the additional two thousand will be especially painful--say, because I have nine thousand dollars more than I "need"--there is no reason for me to take that risk.

I see no way to conclude which one is "more" rational without making assumptions about the subjective preferences of the person making the decision.

This is wrong. On iterated plays you average losses of 8,500 per play, whereas taking the certain loss you average losses of 8,000. 8,000 is better than 8,500.

True. Unless -- http://forums.jolt.co.uk/showpost.php?p=13194662&postcount=21. In which case, my options are 1) Go for the 15% chance Hail Mary, 2) go down in flames, 3) go down in slightly more flames.

Thus, one goes with the 15% chance.

This is true. On an iterated experiment, you average wins of 8,500 per play rather than 8,000 in wins per play. 8,500 is better than 8,000.

But since iteration has been eliminated from possibility ( http://forums.jolt.co.uk/showpost.php?p=13194662&postcount=21 ), is it still true? Even without iteration, throwing away a 85% chance in favor of the best possible outcome is still a bad choice.

You cannot claim to eliminate iteration and simultaneously claim that you have a logically consistent definition of "rational."

If it is "rational" for me to take the bet, then it is always rational for me to take similar bets in similar circumstances; if it is "rational" for me not to take the bet, then it is always rational for me not to take similar bets in similar circumstances.

"Rational" either describes an attitude toward well-defined choices, or it defines nothing objective at all.

Now, I'm not the one claiming "rational" has to describe something objective. But if you are, iteration is part of your model whether you like it or not--or rather, to be more accurate, expected value (not expected utility) is part of the equation. The notion of "iterated experiments" is a convenient, hypothetical way in which to describe the results of attitudes toward taking bets.

Economists don't necessarily understand this... but then, I'm not going to claim economists ever had a consistent (or objective) concept of rationality.

:p

It's not "irrational" in any objective sense--it depends how much I value that additional two thousand dollars compared to how much I fear the risk of losing it all.

If I'm content with 8,000, if I'm not that concerned about the extra, there's no reason to take the risk.


Of course, it is entirely possible that one is perfectly content with gaining and losing nothing, making this game entirely pointless. But that's no fun.

At any rate you can't "risk...losing it all." Presumably, one went into the scenario with x, and at the very worst you're leaving with x (zero gain). That 8,000 gain can't be yours until the scenario is resolved. The ease of access to the 8,000 gain (one need only choose it) is, again, the red herring that leads to the sub-optimal outcome.

But if one happens to roll the 15% chance of no gain, one hasn't "lost" anything, because one never had the 8,000 gain to begin with. One walks away with exactly what one had to begin with.

Judging from the language of the OP, I would anticipate that the ultimate answer to this scenario has something to do with how the psychological need for a guaranteed win ends up causing one to ultimately gain less.


Yes--if the scenario occurs repeatedly. We have no reason to make any such assumption.


Of course.

You cannot claim to eliminate iteration and simultaneously claim that you have a logically consistent definition of "rational."


Who is "you?"

Judging from the language of the OP, I would anticipate that the ultimate answer to this scenario has something to do with how the psychological need for a guaranteed win ends up causing one to ultimately gain less.But it doesn't, at least not always.

Here, a preference for an 85% shot at 10,000 over a guaranteed 8,000 is a preference for (on average) 8,500 over 8,000. It doesn't matter if you never get to make this bet again: what matters is how you evaluate bets in your life.

That means that other bets would work differently. If the choice were a guaranteed 9,000 over an 85% chance of 10,000, the rational person (the person with an attitude maximizing gains) should take the sure thing. Likewise, if the choice were a guaranteed 8,000 over a 75% chance of 10,000, the rational person should take the 8,000--because it represents an attitude that, in the long run, prefers 8,000 to 7,500.

Who is "you?"Anyone trying to measure "rationality" against this test, insisting that iteration is irrelevant. In this case, Cosmopoles.

Of course, it is entirely possible that one is perfectly content with gaining and losing nothing, making this game entirely pointless. But that's no fun.

I thought of that too, but I decided I would keep myself confined to a person conceiving of utility in monetary terms, and being concerned with maximizing it.

At any rate you can't "risk...losing it all." Presumably, one went into the scenario with x, and at the very worst you're leaving with x (zero gain).

Lose all potential rewards, I meant. The reasoning remains the same: if I am content with winning the first eight thousand, if I am not so concerned with the extra two thousand I will get by going with the second option, there is no reason to take the risk of gaining nothing.

But if one happens to roll the 15% chance of no gain, one hasn't "lost" anything, because one never had the 8,000 gain to begin with. One walks away with exactly what one had to begin with.

And 8,000 less than what one could have been guaranteed if one had gone with Option 1.

Judging from the language of the OP, I would anticipate that the ultimate answer to this scenario has something to do with how the psychological need for a guaranteed win ends up causing one to ultimately gain less.

But any such answer depends on the notion that I should care equally about every dollar I acquire--not only that my utility is only defined in monetary terms, which I can accept as a legitimate assumption of the problem, but that each dollar gives me an equal quantity of utility.

Among other things, such an assumption runs into the obstacle of diminishing marginal utility: all else being equal every additional dollar will in fact mean less to me. Enough to compensate for the $500 loss when we consider the average? That depends on my subjective preferences.

In Scenario A, the worst possible outcome is 0. No gain, no loss. The best possible outcome is 10,000. The 8,000 is a red herring because it is irrational to throw away an 85% chance in favor of the best possible outcome for a sub-optimal outcome.

In the long run, you'll make much more money taking the 85% risk then you will playing it safe (assuming the Scenario occurs repeatedly, anyway...)



In Scenario B, the worst possible outcome is -10,000. The best possible outcome is 0 (no loss). The -8000 (first option) is again a red herring. Yes, there is only a 15% chance of getting away with no loss. But by choosing the -8000 loss outright, you are reducing that 15% long shot to 0% yourself.

Choosing to reduce a long shot possibility to a guaranteed loss eliminates outright the only means by which you could get away with no loss. Reducing your options in such a manner as to produce a guaranteed less than optimal outcome seems to be highly irrational.

This is true. On an iterated experiment, you average wins of 8,500 per play rather than 8,000 in wins per play. 8,500 is better than 8,000.

This is wrong. On iterated plays you average losses of 8,500 per play, whereas taking the certain loss you average losses of 8,000. 8,000 is better than 8,500.

Clarification for Loyal Opposition's assumption:

The scenarios are not continuous, i.e. they do not repeat - they're a one off

It's not "irrational" in any objective sense--it depends how much I value that additional two thousand dollars compared to how much I fear the risk of losing it all.

If I'm content with 8,000, if I'm not that concerned about the extra, there's no reason to take the risk.



Yes--if the scenario occurs repeatedly. We have no reason to make any such assumption.



Not at all. Just as with the first case, if I am content with losing the eight thousand but fear that losing the additional two thousand will be especially painful--say, because I have nine thousand dollars more than I "need"--there is no reason for me to take that risk.

I see no way to conclude which one is "more" rational without making assumptions about the subjective preferences of the person making the decision.

True. Unless -- http://forums.jolt.co.uk/showpost.php?p=13194662&postcount=21. In which case, my options are 1) Go for the 15% chance Hail Mary, 2) go down in flames, 3) go down in slightly more flames.

Thus, one goes with the 15% chance.

But since iteration has been eliminated from possibility ( http://forums.jolt.co.uk/showpost.php?p=13194662&postcount=21 ), is it still true? Even without iteration, throwing away a 85% chance in favor of the best possible outcome is still a bad choice.

Of course, it is entirely possible that one is perfectly content with gaining and losing nothing, making this game entirely pointless. But that's no fun.

At any rate you can't "risk...losing it all." Presumably, one went into the scenario with x, and at the very worst you're leaving with x (zero gain). That 8,000 gain can't be yours until the scenario is resolved. The ease of access to the 8,000 gain (one need only choose it) is, again, the red herring that leads to the sub-optimal outcome.

But if one happens to roll the 15% chance of no gain, one hasn't "lost" anything, because one never had the 8,000 gain to begin with. One walks away with exactly what one had to begin with.

Judging from the language of the OP, I would anticipate that the ultimate answer to this scenario has something to do with how the psychological need for a guaranteed win ends up causing one to ultimately gain less.



Of course.

But it doesn't, at least not always.

Here, a preference for an 85% shot at 10,000 over a guaranteed 8,000 is a preference for (on average) 8,500 over 8,000. It doesn't matter if you never get to make this bet again: what matters is how you evaluate bets in your life.

That means that other bets would work differently. If the choice were a guaranteed 9,000 over an 85% chance of 10,000, the rational person (the person with an attitude maximizing gains) should take the sure thing. Likewise, if the choice were a guaranteed 8,000 over a 75% chance of 10,000, the rational person should take the 8,000--because it represents an attitude that, in the long run, prefers 8,000 to 7,500.

I thought of that too, but I decided I would keep myself confined to a person conceiving of utility in monetary terms, and being concerned with maximizing it.



Lose all potential rewards, I meant. The reasoning remains the same: if I am content with winning the first eight thousand, if I am not so concerned with the extra two thousand I will get by going with the second option, there is no reason to take the risk of gaining nothing.



And 8,000 less than what one could have been guaranteed if one had gone with Option 1.



But any such answer depends on the notion that I should care equally about every dollar I acquire--not only that my utility is only defined in monetary terms, which I can accept as a legitimate assumption of the problem, but that each dollar gives me an equal quantity of utility.

Among other things, such an assumption runs into the obstacle of diminishing marginal utility: all else being equal every additional dollar will in fact mean less to me. Enough to compensate for the $500 loss when we consider the average? That depends on my subjective preferences.

"In other news, the Central Intelligence Agency bought a commanding share of Jolt's stock..."

Well, from a statistical standpoint thing to do is go A2, B1. This will (probably) result in maximum gain or minimum loss, depending on which scenario you're in.

However, I'd go A1, B2 because humans (and the Universe) are not rational.
A1 because, if I bet $8,000 on the potential of another $2,000 gained, and then lost, I'd hate myself for (at least) several days. And that assumes that for the purposes of this hypothetical I'm wealthy enough that $8,000 really isn't that big of a deal.
B2 because, I don't plan on paying this debt anyway (the sort of asshole who'd put me in this situation and then decide to try and play games doesn't deserve to get my money) and it would be nice to ditch the tab without having to cut town and change my name.

I would do a barrel roll. Then I would play game 1 and take the $8000. Afterwards, I would spend it all on delicious cake. I wouldn't play game 2 as I spent all my money on delicious cake and can no longer afford to pay either debt.

I would do a barrel roll. Then I would play game 1 and take the $8000. Afterwards, I would spend it all on delicious cake. I wouldn't play game 2 as I spent all my money on delicious cake and can no longer afford to pay either debt.

That seems logical.

But seriously, I chose:

Scenario A - Option 1
Scenario B - Option 1

Why would I need $16,100 (the approximate equivalent of £8.000) anyway?

Enough with the spoilers, too.

"In other news, the Central Intelligence Agency bought a commanding share of Jolt's stock..."

Shhh!

In Scenario A, the worst possible outcome is 0. No gain, no loss.
That is incorrect - the worst outcome is losing 8k. You have forgotten the opportunity costs of not choosing option 1.

Option 1 is effectively nothing - you can take it as the baseline.
Option 2 is therefore a 15% chance of losing 8k, or an 85% chance of gaining 2k.

When you put it in those terms, a person can calculate the marginal benefit of that 2k gain as being significantly lower than the marginal cost of an 8k loss, despite the difference in probabilities. Or if factoring in the probabilities, a loss of 1200 outweighs a gain of 1700 in terms of marginal utility (it all depends on how you draw the curve).


Never forget opportunity costs - or in other words, if someone offers you a choice between a candy bar and an identical candy bar, they are effectively offering you nothing (even economics is funny at times).

Further proof offered here:
http://www.youtube.com/watch?v=VVp8UGjECt4

Further proof offered here:
http://www.youtube.com/watch?v=VVp8UGjECt4

That was very funny - thanks!

But any such answer depends on the notion that I should care equally about every dollar I acquire--not only that my utility is only defined in monetary terms, which I can accept as a legitimate assumption of the problem, but that each dollar gives me an equal quantity of utility.

Among other things, such an assumption runs into the obstacle of diminishing marginal utility: all else being equal every additional dollar will in fact mean less to me. Enough to compensate for the $500 loss when we consider the average? That depends on my subjective preferences.
Yeah, you hit the nail on the head there. For example in my current situation I can't afford a car but gaining either £8000 or £10000 would change that. Which of the amounts I get to spend on the car is only an insignificant minor detail compared to the fact of being able to afford a car at all. In this situation the first £8000 are worth so much more then the additional £2000 to me so I'd play it safe even though game theory says £10000*0.85 > £8000.

That is incorrect - the worst outcome is losing 8k. You have forgotten the opportunity costs of not choosing option 1.

Option 1 is effectively nothing - you can take it as the baseline.
Option 2 is therefore a 15% chance of losing 8k, or an 85% chance of gaining 2k.That's not how opportunity cost works. Or, if you choose to do the calculation that way, you have to see what happens when you make the other choice the baseline instead.

If I choose Option 1 my net gain is my gain (8k) minus (-) my opportunity cost in not choosing Option 2. But the expected value for Option 2 is 8.5k, so my net gain taking Option 1 is -.5k.

If I choose Option 2 my net gain is 8.5k - 8k = .5k.

Of course, this again presumes a rationality that does not exist for one moment of my life; rather, it guides the kinds of risks I should be willing to take. Otherwise, how am I to know when it is better to gamble?

Note that the result according to YOUR calculation is identical when the probability is 90% to win, or 95% to win. Either you have to pick some arbitrary point at which it is worth "risking" your 8k opportunity cost to gain 2k (99%? 99.9?), or you have to admit that your analysis always favors taking the "sure thing." In the first case, if the tipping point is arbitrary (or subjective, which amounts to the same thing), then you have to admit that you do not have a consistent definition of "rational" and you can't "test" anything here... while in the second case, I think you're proposing a definition of "rational" with which few rational people would agree: most of us think there are SOME "good bets."

There seems to be two camps right now, with one group claiming that A1B1 is rational and another claiming A2B2 is rational. Both groups have some very good arguments as to why each is correct, and with good reason too - both are rational!

This is known as Prospect theory (http://en.wikipedia.org/wiki/Prospect_theory). The key to this problem is that both scenarios are essentialy the same - the only difference is whether your money goes up or down. Therefore, a rational person would choose option 1 or option 2 in both scenarios. The only difference is how averse to risk you are. If you are quite risk averse you will choose options 1, rather than risking higher gains or losses on a gamble. However, if you are risk neutral, or even risk loving, you will take options 2, and gamble both times. It all comes down to personal preference.

An irrational person will normally choose A1B2. As the poll shows, and the original study showed, the majority of people go for this option. Its known as the endowment effect where people value something they already have higher than something that is exactly the same that they could potentially win.

And so, once again, we have evidence that people are not rational economic actors.

Someone should inform the libertarians...

And so, once again, we have evidence that people are not rational economic actors.

Someone should inform the libertarians...

In standard finance it was assumed that although people are irrational, there's enough rational people to effectively cancel out their effects despite there being less rational people.

Behavioural finance is more popular (and has been since about the mid 90s) and says that irrational people do have a noticeable effect.

I always figured my risk-averse behavior was justified. It simply doesn't make sense to me to gamble away a guaranteed return for the change of a larger one that is nonetheless smaller than the potential loss. I also feel it's better to lose a guaranteed 8,000 than gamble and end up losing 10,000 for the low chance of losing nothing.

A1B2.

To me, the sure 8k, is better than a chance at only 2k more. Even if its a good chance.

However, to me, the chance at losing nothing is worth taking, as the total loss (assuming I lose the full 10k) is only 2k more.

Why?

I dont know.

A1B2.Why?

I dont know.

Most people would say the same.

An example of this effect in real life would be if a person has shares in two companies, X and Y. The value of shares in X has been increasing while shares in Y have been decreasing. This means that the typical (irrational) investor will sell his shares in X (rather than risk them falling) while holding on to shares in Y (they've already fallen, so I'll just see if they fall further...). This defies conventional asset market wisdom, which says you hold on to rising assets and ditch falling assets.

I'd rather just have the 8000. I don't care if it's supposedly "irrational." If I go for the 10,000, I can still get nothing. If I take the 8000, I don't have the most possible money, but it's still a lot of money. The extra 2000 really isn't all that meaningful to me in the face of how stupid I'd feel if I bet on it and lost.

For the second question, I choose the second option. Again, the extra 2000 isn't all that important to me, since 8000 is by far a bigger number.

Most people would say the same.

An example of this effect in real life would be if a person has shares in two companies, X and Y. The value of shares in X has been increasing while shares in Y have been decreasing. This means that the typical (irrational) investor will sell his shares in X (rather than risk them falling) while holding on to shares in Y (they've already fallen, so I'll just see if they fall further...). This defies conventional asset market wisdom, which says you hold on to rising assets and ditch falling assets.

So...yay! Im typical?

An irrational person will normally choose A1B2.
In my financial situation if I get 8000, I regain full financial fluidity (it's marginal now). If I lose 8 or 10 thousand, I'm screwed. Tell me now my choice wasn't rational.
IOW, you have made some unwarranted assumptions.

Then why not take the gamble to gain 10000?

85% of 10000 is 8500, so on average you have a better chance of winning or losing more money.

Therefore, A 2, B 1

Given only option B and my current financial situation, I would choose B 2.

Then why not take the gamble to gain 10000?

Because you could lose all of it. It's better to have 8,000 guaranteed than to gamble it all on a potential that might end up costing you everything. To quote than ancient saying, a bird in the hand is worth two in the bush.

Then why not take the gamble to gain 10000?
If you are referring to me, then because I value full financial fluidity higher than 2000 (or rather 1700). As it stands now, if i have any unexpected expenses, I'm screwed. So you may say that I'm extremally risk averse when it comes to gambling my financial fluidity away ;).

There seems to be two camps right now, with one group claiming that A1B1 is rational and another claiming A2B2 is rational. Both groups have some very good arguments as to why each is correct, and with good reason too - both are rational!You need to check your numbers.

This is known as Prospect theory (http://en.wikipedia.org/wiki/Prospect_theory).No. No, it's not.

Prospect theory would predict A2B1, because prospect theory predicts that people are risk-seeking with "found" money (scenario A) but risk-aversive with their "own" money (Scenario B).

No. No, it's not.

Prospect theory would predict A2B1, because prospect theory predicts that people are risk-seeking with "found" money (scenario A) but risk-aversive with their "own" money (Scenario B).

I'm 100% certain about this. Read it here (http://www.econport.org/econport/request?page=man_ru_advanced_prospect). "Simply put - while they are risk-averse over prospects involving gains, people become risk-loving over prospects involving losses."

In Scenario A, the worst possible outcome is 0. No gain, no loss. The best possible outcome is 10,000. The 8,000 is a red herring because it is irrational to throw away an 85% chance in favor of the best possible outcome for a sub-optimal outcome.

In the long run, you'll make much more money taking the 85% risk then you will playing it safe (assuming the Scenario occurs repeatedly, anyway...)




Yes, in a repeated scenario, it's true. But in a one-shot choice, it's not having 8500 in average or having 8000 in average, but being sure to have 8000 or taking a risk to have nothing.

What matters is not to maximize money, but to maximize happiness (for me, my friends/relatives and/or foreign people, because I can also give away part of the money, I'm that kind of person). But "happiness" is not proportionate to money. The improvement of my life by earning 10000 is not 25% above the improvement granted by the 8000. The more money you already have, usually, the more money you need for the same improvement in living conditions/happiness. So in this case, I would chose the 8000.


In Scenario B, the worst possible outcome is -10,000. The best possible outcome is 0 (no loss). The -8000 (first option) is again a red herring. Yes, there is only a 15% chance of getting away with no loss. But by choosing the -8000 loss outright, you are reducing that 15% long shot to 0% yourself.

Choosing to reduce a long shot possibility to a guaranteed loss eliminates outright the only means by which you could get away with no loss. Reducing your options in such a manner as to produce a guaranteed less than optimal outcome seems to be highly irrational.

In this case, the average loss in 8500 for option B and 8000 for option A, so even purely rationally, option A is better. But the main argument of the first case still holds: improvement granted by money is not proportional to it. Depending on how much money you have/earn abd what are your needs, losing 10k can be twice as hard as losing 8k, because it may needs starting to give up things really important.

Wow. I don't know what compelled me to choose B2 now...

Good job it is in pounds, if it was dollars by the time you won the lottery the prize could be almost worthless.:p

For the people that say that their decision was made based on things like "with 8K I can buy a car already so I don't need the possible extra 2K" what if the numbers were just 8 and 10 pounds? This would (hopefully) cripple noone financially, thus making those arguments less relevant.

The key to this problem is that both scenarios are essentialy the same - the only difference is whether your money goes up or down.

Yeah, and that's a very big difference.

For starters, the average result from choosing Option 2 in the first case is +$8,500, compared to +$8,000 for Option 1. The average result from choosing Option 2 in the second case is -$8,500 for Option 2, compared to -$8000 for Option 1. This means that if I am risk-neutral and my economic situation is such that the utility I gain from each dollar is more or less equal, I will go for Option 2 in the first case (because, on average, I will gain $8,500 as opposed to $8,000) and Option 1 in the second (because, on average, I will lose $8000 as opposed to $8,500.)

I'm 100% certain about this. Read it here (http://www.econport.org/econport/request?page=man_ru_advanced_prospect). "Simply put - while they are risk-averse over prospects involving gains, people become risk-loving over prospects involving losses."

That is hardly unexpected when one considers that rational economic actors maximize their own satisfaction, not some objective amount, and that economic actors tend to abide by diminishing marginal utility, especially when dealing with money and easily managed amounts.

85% of 10000 is 8500, so on average you have a better chance of winning or losing more money.


But you get only one shot. I would just assume that the thing with 85% likelihood was going to happen and base my decision on that.

Just to add my thoughts on this...

Supposing at the start of both scenarios, I am already in possession of $10,000 and that is all the money I have to my name.

If a random stranger came up to me and offered me $8,000 or a game of chance where I have pretty good odds of walking away with $10,000 if I win, or nothing at all if I lose, I'm going to take the $8,000 simply because I perfer the certainty of the gain. $2,000 more doesn't mean that much to me when I'm walking away with $8,000.

Conversely, assuming again that I am in possession of only $10,000 to begin with, in scenario B, let's suppose some twisted criminal came up to me and put a gun to my head and demanded I give up $8,000 of my $10,000 or, play a game of chance where I have a minimal chance of losing nothing at all and a very likely chance that I will lose everything. I'd be more inclined to give over my $8,000 to the crazy person and be damn glad that I still have $2,000 at all than to chance walking away penniless.

It's how I see it, anyway.

For the people that say that their decision was made based on things like "with 8K I can buy a car already so I don't need the possible extra 2K" what if the numbers were just 8 and 10 pounds? This would (hopefully) cripple noone financially, thus making those arguments less relevant.

I'd still take the certain gain over the chance that I could waste my time and come away with nothing to show for it. I tend to have shitty luck.

said A1, B1. What's the "real" answer.

If the scenarios occurred frequently, I'd go with A2 and B1; as it is, I'd go with A1 and B2.
Expected outcome is not the only measure of rationality. Besides which, the utility of 10000 is only marginally more than 8000; the marginal increase is not worth risking 8000 at 15% probability. And in the other scenario, it plays out similarly. Not losing 8000 is worth risking another 2000 (because that's only a marginal increase in utility).

said A1, B1. What's the "real" answer.I would hazard to guess that what has been bombarded to "rational" status would be A2 and B1.

Whether it is, depends on whether that strategy would increase you reproductive success in life. (From an evolutionary standpoint of rationality).
If (sc. B) losing 8000 means you end up dead (because you can't pay off mobsters you're indebted to), you may as well opt for that 15% chance to live. If likewise (sc. A) you need only get 8000 to pay off the mobsters out to kill you, well, it would be rather irresponsible to risk your life to get ahead 2000.
It's all about context. Different background stories can just as easily turn things around.

I was writing an essay on behavioural finance the other week - which, if you're interested, is the use of psychology to try and predict how investors behave and then use the behaviour to make the most of investments - and I came across an economic theory which described a test to judge economic rationality. Economic rationality of investors is one of the assumptions (to a certain degree) that standard finance is based on, and behavioural finance challenges this by claiming that investors are irrational.

In this test there are two scenarios. Scenario A:

You have a choice of two options. In option 1, you recieve £8000 guaranteed. In option 2, you enter a lottery, with 85% chance of winning £10,000 and 15% chance of winning nothing.

Scenraio B:

You have a choice of two options. In option 1, you lose £8000 guaranteed. In option 2, you enter a lottery, with 15% chance of losing nothing and 85% chance of losing £10,000.

So which options from each scenario would you choose? Try to make your decision without looking up the economically 'rational' answer. If anyone's familiar with the theory, please don't spoil it until we hear what people have to say.


Scenario A: Option1 and Secnario B: Option2

Why choose differently: because of the amount involved. That's quite much money. If exercise would say 80 and 100 pounds with possibility that such cases would repeat over time, my choice would be otherwise. 8000 is much and I prefer to get them (I'm being risk averse investor in this time). In Scenario B since i do not have 8000, then Option1 for me spells 'loosing everything' and Option2 is '85% probability of loosing everything and 15% of loosing nothing'.

This is known as Prospect theory (http://en.wikipedia.org/wiki/Prospect_theory). The key to this problem is that both scenarios are essentialy the same - the only difference is whether your money goes up or down. Therefore, a rational person would choose option 1 or option 2 in both scenarios. The only difference is how averse to risk you are. If you are quite risk averse you will choose options 1, rather than risking higher gains or losses on a gamble. However, if you are risk neutral, or even risk loving, you will take options 2, and gamble both times. It all comes down to personal preference.

An irrational person will normally choose A1B2. As the poll shows, and the original study showed, the majority of people go for this option. Its known as the endowment effect where people value something they already have higher than something that is exactly the same that they could potentially win.

Well, well. There is difference (one scenario is receiving and other is losing money) so this difference surely influence the choice. Whether these reasons are irrational, it doesn't tell. They may be rationally influenced by other decisions (level of wealth, investment portfolios). I would suggest testing this theory on groups with different levels of wealth, and providing different amounts of losing/gaining money.