Ergodicity for Silicon Valley

Ergodicity is derived from Ergodic Theory, which is the study of the long-term average behavior of systems evolving in time. Ergodicity describes a system where the time average of any property is equal to the ensemble average of that property. In other words, it means that the long-term behavior of a single system is equivalent to the average behavior of many identical systems at a given time.

While ergodicity is used in physics and mathematics, the idea can be applied more broadly in other fields but remains underrated. For the purposes of this post, I will provide a simple overview of ergodicity and focus more on real-life examples & its implications.

We often assume that most systems are ergodic. But, in reality, nearly all human systems are non-ergodic. Misinterpreting non-ergodic systems as ergodic is a terrible ‘category error’ made by not only mainstream economists but also by corporate leaders and entrepreneurs. Even in personal life, understanding ergodicity is crucial to avoid catastrophic outcomes like bankruptcy. Let’s dive in…

Table of Contents

• Will you take this bet?

• Leverage in Investing

• ‘Sequence of Returns’ Risk

• Loss Aversion is not a bias

• Why I’ll never become a restauranteur

• Cybersecurity

  • Adams’ 3rd Law

  • What if everyone did it?

• Corporate Leaders

  • Mean is Meaningless

  • Customer Obsession

• Entrepreneurs

  • Why I hate Y combinator?

  • Take some chips off the table

Will you take this bet?

Ole Peters is the pioneer of Ergodicity Economics. He explains how the standard ‘expected utility’ toolbox for evaluating wagers is a flawed basis for rational decision-making, with a simple game of betting…

You start with a $100 balance. You flip a coin. If it lands on heads, you win 50% of your current balance. If it lands on tails, you lose 40%. Then you repeat the process - this is the game. Will you take this bet?

Here, the ‘expected value’ (based on the ensemble average) of the first coin toss is (0.5*$50 + 0.5*$-40) = $5, or a 5% gain. Using this insight, the expectation for the second sequential coin toss is (0.5*52.5 + 0.5 * $-42) = $5.25, another 5% gain.

Intuitively, it seems like a good bet and that the house will lose money as the odds of heads or tails is 50% each. But, if you are wrong, you only lose 40% of your money in any toss.

Here’s why it is likely you’ll go bust playing this game. Say you get 2 tails in a sequence, you are already down to $36. Now, you need 3 consecutive heads to get above where you started.

What matters in real-life, is the time-series outcome (which is multiplicative) not the ensemble average. The average outcome of thousands of people playing the game at a single time doesn’t matter to me. I carry my own ‘sequence of returns’ risk.

Peters asserts that the smart approach is to focus on maximizing the average expected rate of growth of the balance over time, rather than the average outcome across various alternatives. If you run a simulation, the average rate of growth over time for all runs is actually - 5.03%. This indicates that it is not a good bet to engage in, despite the mainstream assessment suggesting an expected return of 5%.

Leverage in Investing

Let’s now apply the concept of ergodicity to investing. In an old post from July (Advice to my 18 year old self), I wrote about the mistake of “leverage” (in investing):

As it turns out, the activist investor Carl Icahn almost blew up because of the use of leverage. He recently settled with the SEC for failing to disclose personal loans secured by his shares in Icahn Enterprises — but that borrowing cost him much more than the $2 million dollar fine. While the numbers are not public, some estimate that his wealth went from $24 billion to less than $4 billion as the shares tumbled from the Hindenburg reporting.

He was forced to put up additional collateral against the loans to avoid margin calls that might have forced a fire sale of his assets. Wall Street is shocked to see him lose >80% of his net worth (at the ripe old of 88) as uncertainty prevails about his financial future.

Life is path dependent. As Charlie Munger elegantly put it:

‘Sequence of Returns’ Risk

Many retirees retire with a naive assumption that they can withdraw 4% of their portfolio every year as the S&P 500 index fund would give them an “average” 7% yearly returns. Some personal finance “gurus” even recommend withdrawing > 6% assuming 11% or 12% “average” returns. Ignoring the fact that past performance doesn’t guarantee future returns, we must note that the sequence of those returns matter a lot. As Nassim Taleb reminds us, you can drown in a river with an “average” depth of 4 feet.

Say in the 2 years after you retire, there is a 50% gain in your portfolio followed by a 50% loss. Now, the average return is 0% and you are back to where you started. But, if the order was reversed, your initial portfolio would be down by 25% even though the “average” return is still is 0%!

Wealth is multiplicative, not additive! You’d be better off by focusing on the time-series returns and not try to squeeze the most returns in any given year. In case you wanna learn how, checkout my previous post (multidisciplinary lens on investing) here.

Loss Aversion is not a bias

Evolution has coded into us certain insights for decision-making as emotions, intuitions or feelings - mainstream calls them “biases”. Not all of them are useful in the modern world - but some are. Loss Aversion is a popular one. Behavioral economics often interprets loss aversion as irrational. That’s because it doesn't optimize over time but instead takes expectations and optimizes over a statistical ensemble.

Stock market returns must never be seen linearly one year at a time. What matters is the cumulative result. That’s how compounding works. A 50% loss in your portfolio needs a 100% gain to recover. Learn to see things through time.

It is widely known that many psychology experiments and research papers don’t replicate. But, I’d be willing to support the hypothesis that papers that offer good explanations from the lens of evolution and ergodicity have a much better chance of getting replicated.

Why I’ll never become a restaurateur

It was Nassim Taleb that got me hooked with the topic of ergodicity. In his ‘Incerto’ series, he discusses about ergodicity. He notes how the restaurant industry as a whole has become so successful over the decades. But, many individual restaurants have died out in the process and existing ones face evolutionary elimination at any given point in time.

I’ve gotten a few proposals and offers to invest in restaurants over the years. But, this insight from Taleb has prevented me from jumping into that business. I have closely kept track of what happened to those investments and they all have gone bust for varying reasons - head chef quit after a competitor hired him away, the property owner drastically raised rent after initial success, global pandemic, etc.

What’s good for the collective may not be good for the individual. Hat’s off to all restaurateurs - I for one, will remain a foodie and support good restaurants as their customer. But, I don’t think I’ll ever venture into the restaurant business.

Use the lens of ergodicity to make better maps of the world we live in… Don’t be fooled by what is good for the collective as it may not be good for you personally.

For the premium-tier subscribers, I’ll now explore the implications of ergodicity on the field of cybersecurity, leadership and entrepreneurship.

Cybersecurity: We’ll take the lens of ergodicity with two examples…
1. The overall strategy/approach and organizational design of cybersecurity teams can negatively impact the whole organization when ergodicity is ignored.
2. Is implementing SMS based One-Time-Passwords (OTP) good or bad for protecting your customers?

Corporate Leaders: Many corporate metrics are ensemble averages. We discuss why ‘mean’ is meaningless for many metrics like Employee/Developer Happiness Survey and the importance of looking at the sequence of customer interactions over time rather than point-in-time ensemble averages.

Entrepreneurs: Here, we’ll apply the lens of ergodicity before you become an entrepreneur (starting from the aspect of funding a new business), during (as you run & grow the business) and after your business becomes successful and discuss key insights applicable for each stage.

Subscribe to "I'm Serious" to read the rest.