The Year of Levered Conviction

My investing journey, Part 6: 2020

Sep 28, 2026 · @Deniz Erkan

In 2020, my portfolio rose 858%. Everyone who sees that number asks the same question, and they should. A return like that either means someone took risks they did not understand, or did something specific that can be explained. Here is exactly what I did.

Options, learned slowly

I had been studying options since 2009, when Buffett's put options first changed how I thought about them. For nearly ten years it stayed theory. At the beginning of 2018, I started using them, increasingly through 2018 and 2019, alongside the stocks I had the most conviction in. By the time Covid hit, I was very well versed in them.

What I did in the crash

Between February 19 and March 23, 2020, the S&P 500 fell 34%. Most investors were trying to get out. I did the opposite, in a specific way.

The main thing I did was sell put options on my highest-conviction companies, such as AMD and Nvidia. A sold put pays you a premium today in exchange for a promise: if the stock falls to a set price, you buy it at that price. I believed in these companies for the long term. If the crash kept going and I was assigned the shares, I would own companies I wanted to own anyway, at prices better than I had ever expected to pay. If it did not, I kept the premium.

For upside, I mainly used long calls on the companies I believed in most. A long call has a known maximum loss: the price you pay for it. Before every position, I asked one question: how much am I willing to lose? I sized the calls by that number, and I sized the puts by the price at which I would be comfortable owning the stock and how much of it I was prepared to buy. I never sold an option I was not prepared to honor.

Why the number is so high

The market recovered from its March low faster than almost anyone expected, and the platform companies I believed in led the way: Roku, Square and JD.com carried most of my year. Options multiply the result of being right. My conviction was levered, so the recovery was levered too. That, not an algorithm and not a secret, is why 2020 looks the way it does.

Take 2020 out

The fair question is whether anything is left without that one year. So here is everything since January 2021, including the worst year of my life as an investor.

Reconstructed time-weighted returns; Nasdaq-100 and S&P 500 price indexes (Yahoo Finance, FRED)

Even with 2020 removed and my worst year left in, I have compounded faster than the technology index. This is the record since my style was set. The years before it, when I had no style and then the flat years, trailed the market, and I have said so in this series.

The other side of leverage

Leverage has no opinion about whether you are right. Two years after 2020, the same approach that multiplied my gains multiplied my losses. In 2022 I lost 47% while the Nasdaq-100 lost 33%, as I described in the previous letter. 2020 and 2022 are not two different stories. They are the same decision, seen in a good year and in a terrible one.

That is why I sized every position by what I was willing to lose, and why that question matters more to me today than it did then. Leverage does not change what you know. It changes how much it costs you to be wrong.

Where Runtime's ideas came from

The record in this series combines all of my personal accounts. Most of them I ran entirely on my own judgment. In one, I had started running early Runtime ideas alongside my own positions, but in 2020 that work was still an experiment, and the year's result came from my conviction positions, not an algorithm. These years of using options inside old-school investing planted the idea that later became Runtime: you can shape the risk and return profile of owning stocks. Take the emotion out, take the human guesswork out, express the positions through rules, and you can decide in advance how much a portfolio should be able to lose and how it should behave when markets break. I learned that by doing it by hand. Runtime is my attempt to do it with discipline no human can sustain alone.

Leverage does not create conviction. It multiplies whatever conviction you already have, including the mistakes.

Next in the series

Part 7, "Borrowed Conviction," is about the ideas that were never really mine, and how they cost me.

Engineered compounding,
Deniz Erkan

Disclosure: This letter describes my personal investment history in my own accounts, not the Runtime fund or its algorithmic strategy. Returns are reconstructed estimates from monthly account balances, are not audited, and do not reflect fees or taxes any other investor would pay. Index comparisons use price returns, excluding dividends. Options involve substantial risk, including the loss of the entire premium paid or losses well beyond the premium received. Past performance does not predict future results. This is not investment advice.