Ergodicity: Contextualizing Risk for the Individual
Originally sent to clients of Upstart Wealth Management
There’s no return without risk. The relationship between the two is well-studied. But the language typically used to describe risk & return applies to populations, not individuals. Academic answers to risk/return tradeoffs apply at the macro level, but are a poor abstraction for the context needed to make important personal decisions. To better equip ourselves with tools to consider risk at the individual level, we connected with Luca Dellanna, author of Ergodicity. This is the first of a series that offers startup/tech ecosystem applications to some of the concepts that Luca shared.
One of the most important takeaways from Luca's work is that individuals should not engage with risk as an average, but as a distribution. As an example, suppose you can put $1,000 into one of two bets, both of which have a 10% expected return. Do you have a preference for which bet you place? Knowing that the bet returns an average of 10% doesn’t tell you much. Instead, you need to know the distribution of possible returns. Suppose for one bet, 50% of bettors lose their money, 38% get their money back, and 12% receive a 5X return; that’s a 10% return on average. For the second bet, 100% of bettors receive a 10% return. Two identical expected returns, two very different real experiences.
Application: Think in terms of distributions, not averages. Remember that a positive expected return doesn’t guarantee a positive outcome; if there was no risk, there would be no return.
The next important concept is that risk/reward decisions don’t happen in a vacuum. Rather, an individual’s specific situation must be taken into account. Let’s consider two people who are both negotiating employment offers with a seed-stage startup:
Cyndi is single, rents an apartment under her budget, and had a successful exit from her last company.
Ryan is married with two young children and has a mortgage that takes up 50% of the family’s after-tax pay
Both are able to choose the proportion of restricted stock vs. cash compensation they receive. They can choose 100% stock, 100% cash, or anything in between. Both are bullish on the stock’s potential returns, but understand there’s plenty of risk as well; distributions could range from the stock being worthless to 50X today’s value. Cash compensation obviously has a much narrower range of outcomes.
Cyndi is able to take on much more risk (higher risk capacity) because of her financial position at this point in time. She prefers positive outcomes, but she won’t be knocked off course if the company goes bust. Cyndi can accept any potential outcome along the distribution from worthless to unicorn. Her decision can be based on her risk tolerance, as her risk capacity means she’ll ultimately be ok regardless of outcome.
Ryan, on the other hand, cannot absorb the downside risk of the stock ending up worthless - such a result could knock him off trajectory. It’s prudent for him to take a larger chunk of compensation in cash, even if the higher expected return of the stock is attractive. His decision is driven by his initial position and limited capacity to take risk rather than the risk he feels he can tolerate.
Application: On paper, we’re often asked to weigh risk vs. return. However, in reality, an important variable is missing: resiliency. Without the resiliency to weather downside risk, choices are informed by risk capacity more than risk tolerance. If someone wants to take more risks, a good first step is to increase their capacity for risk. Do this by either increasing resources or sharing risk where possible, like insurance coverage. Also consider isolating certain resources from your riskiest bets, so if you lose your shirt, you at least keep your pants and shoes.
The last concept is that risks are often not experienced in isolation, but in series.
Let’s consider a hypothetical investment opportunity with the following structure:
Each year, the fund buys a stake in one distressed company. At the end of the year, they sell their stake. Next, they buy into another distressed company with the previous year's proceeds. They fund has a 25% expected return. However, there’s a 15% chance the company they buy goes bankrupt in any given year and the fund closes down.
Question: after 5 years in the fund, what are the odds the investment went bust? The 15% risk of bankruptcy in any given year doesn’t happen in isolation; those risks stack together:
By the end of year 5, the chance the fund went bust is 56%!
If this fund had a 5-year lockup (you can’t take money out for 5 years), it would be a hard bet to stomach. If you could pull your funds out at the end of any given year, you may be inclined to roll the dice for a single year. Taking on risk in isolation is very different from maintaining exposure to the same risk over time, especially one with go-to-zero potential.
Application: Risks are not isolated, they compound over time. “Letting it ride” can lead to huge winnings, but it can also erase all previous gains in a flash. If you’re willing to redistribute from your risky winners to safer bets, you increase your resiliency. This, in turn, allows you to keep placing risky bets with a portion of your wealth, knowing you’ve limited your exposure to system-wide blowup risk.
