What is risk?

Ok, so you're a gazelle. And you're hungry. But every time you put your head down to munch some grass you take your eyes off the horizon and for sure there's a lion out there somewhere that wants to get you. But you don't know how likely that is. 

Now you're the lion. Every time you attempt to catch a gazelle it takes a lot of energy. If you fail, that's energy wasted. If you fail enough times you'll be so hungry that you won't be able to get one at all and you might starve. 

Everybody has to make decisions under uncertainty. And every species has figured out behaviours that balance risk and reward such that the individuals of the species leave more offspring. 

Here, I'm using the anthropomorphizing language of what a species wants or figures out, though in reality some individuals just by chance have a different relationship to risk that happens to confer a competitive advantage...they leave more offspring... And if the behaviour is heritable then their descendents will also share this risk profile.

Gazelles and lions think about risk differently. A gazelle has food everywhere. Stopping for a sec to scan the horizon is not a big cost, there's still grass there and it just missed a few moments of munching. However, if the gazelle doesn't see the lion hiding in the bushes, it's a big cost. So gazelles generally spook easy. If there's a risk of a lion... they just run. Maybe it was a lion or maybe just a scary looking branch. Gazelles are pretty fearful.

For a lion, there isn't food lying around everywhere. Food takes effort. Every time a lion goes to get food it is a big investment for a big payoff. Lions have to be much more comfortable with risk. There are lots of options around but chasing after every gazelle greedily is probably a bad idea. Lions have to manage their greed.

Humans are kind of in the middle. Some fear is good, some greed is good. Too much of either is bad.

We like to think of ourselves as perfectly rational (or at least, in the Renaissance we did). Even up to the birth of game theory (von Neumann in the 40s, and then Iterated Prisoner's Dilemma investigations during the Cold War) we were pretty interested in proving humans were rational. 

Turns out we're not rational, and evolutionary game theory (John Maynard Smith, 1973) and third waves feminism helped us come to terms with it. What we do is a pretty good job approximating rationality. Our thinking wasn't designed, it evolved. Our relationship to risk evolved to solve the kind of problems humans had to face. A bunch of weird problems... We're omnivores with crappy claws and advanced social structure that facilitate complex tool making/use and food gathering strategies.

We even share some of our cool irrationalities with chimps. We are both loss averse, meaning that if we have a banana we care more about that banana than a banana lying around that's not currently in our possession. We've done economics experiments with chimps to show it. We even teach them about a unit of currency that they can exchange for food and display a remarkably similar irrationality to our own. 

(Remarkably, as soon as chimps understood the meaning of currency they invented sex work).

Our relationship to risk also depends on our gender and age. Men as they enter breeding age are incentivized to take big risks to secure a mate. This was true of our ancestors that fought each other for mates and young men now drive around real fast instead. Maybe they think this will help secure a mate, though their target demographic seems to disagree with it as a successful strategy (at least, that is my anecdotal evidence).

Risk also helps us understand the boundaries of what is safe. A kid who experiments with running early will learn balance faster, but expose themself to higher risk of falling. Interestingly, in cultures where toddlers falling over carries higher risk (say around open fires or precipitous falls), the average age of walking is often delayed. The risk reward equation changes, and so to does the evolved answer.

So unfortunately, our relationship to risk evolved to solve food and reproductive challenges and not to navigate a complex mathematical system like an economy. 

On the upside, our ability to reflect on our emotional reality allows us to change our emotional reality. We can, through our curious brains, change our relationship to risk.

Pause. What does risk mean to you? Which parts of this resonate, which parts missed?

What do we mean by risk, mathematically?

What would you prefer? (i) \$100; or (ii) a 1 in 10 chance to win \$1100?

They are fairly similar on average, with option (ii) being a little higher. (ii) is also riskier. What do we mean by this?

Well 1 out of every 10 times in option (ii) we get \$1100, and in 9 out of 10 times we get \$0. The average then is: $avg = (1/10)*\$1100 +(9/10)*\$0 = \$110$

Have you encountered averages before? How do you conceive of an average?

And what do we mean by risk? Well we sort of mean... how often do we get an outcome different than the average, and how different is it than the average? Another word for average is expected value. The variance is how different our experience is from expectation.

Since the difference between expectation and reality can be positive or negative, let's just take the difference squared. ie. when we do win the \$1100 it's a lot bigger than average $1100-110 = 990$, and when you don't win the \$1100 you're $0-110 = -110$ less than the average. So we square these numbers so we're measuring the magnitude of the difference, rather than caring about if it's positive or negative. Then, we just take an average of these numbers

9 out of 10 times we are $\$-110$ below average (square this number: $(-110)^2$) and 1 out of 10 times we are $\$990$ above average (square this number: $990^2$. Add these together with the weights of how opften they happen and we get: $var = 0.1*(110^2)+0.9*(990^2) = 108900$. This is kind of an annoyingly big number (because we squared the numbers before adding them) so let's take it's square root. We call that the standard deviation, and in this case: $stdev = 330$. That's a good size of number :)

So option (ii) has a higher average but also a higher variance. Option (i) has a lower average but no variance.

Which one would you take?

What if, instead of a $\$1100$ payoff for the risky option it was $\$10000$? 

How big would the payoff have to be for you be equally happy with the options?

Does it depend on how much money you have currently?

Would you have answered differently 10 years ago? What about 10 years in the future?

These are feelings exploration questions to help guide you in an investigation of your relationship with risk. There are no right answers.

(i) Risk neutral would mean that imyou take the higher average option.

(ii) Risk averse means you are willing to take a lower average in order to reduce the variance you experience (most humans fall here).

(iii) Risk loving means having an element of risk activity makes you more attracted. You would take a lower average payout if it was riskier.

Which one resonates with you?

Let's play some games!

Game 1 - play the above game 20 times
(i) intro
(ii) play once
(iii) compare answers
(iv) explain

Game 2 - 2/3 of the average
(i) intro
(ii) play once
(iii) compare answers, explain
(iv) play again


Game 3 - iterated prisoners dilemma
(i) introduce prisoners dilemma. [C-C = 1,1; C-D = -1, 2; D-D = 0]
(ii) provide currency - everyone gets 100 dice.
(iii) pair off, play iterated prisoners dilemma. bank provides extra dice after each round
(iv) do all the pairs
(v) count dice at the end 

This post is a little long, so we'll talk about risk tolerance in investing in a subsequent post. But next we need to know what a stock is.

Thanks for reading!



Comments

Popular posts from this blog

What is an RRSP, TFSA, FHSA, RESP?

What is debt?