Borrowed Conviction
My investing journey, Part 7
Sep 28, 2026 · @Deniz Erkan
The previous parts of this series were about conviction I earned: the companies I understood because I had built products on top of them. This one is about the other kind.
Some of my investments were never really mine. The idea came from someone I respected, or from an analogy that felt right, and I bought it without building the understanding myself. I call that borrowed conviction. It feels identical to the real thing on the day you buy. It behaves very differently on the day the stock falls.
Ideas from friends
Almost every small biotech company I ever owned came from friends whose expertise I deeply respect. T2 Biosystems, which makes diagnostics that detect the causes of sepsis quickly, was one. Aclaris Therapeutics, an immunology drug developer, was another. My friends knew their science far better than I did, and their reasoning was sound.
The outcomes were brutal. T2 Biosystems went through two reverse stock splits totalling 5,000 to one and left the Nasdaq in 2025; measured from its 2018 high, almost nothing is left. Aclaris fell 98% from its 2021 high to its 2023 low. My own positions were small and I was out of both before the worst of it, so the money I lost was modest. The lesson was not.
The problem was not my friends' judgment. It was mine. When the news turned against these companies, I could not tell whether a thesis was broken or just delayed, because I had never owned the thesis in the first place. If I cannot explain why I own something without quoting someone else, I do not really own it.
Conviction borrowed from an analogy
Farfetch was my largest single company loss, and it was borrowed in a subtler way. Nobody sold me on it. I sold myself, with an analogy.
I thought of Farfetch as the cloud of high-end clothing: a technology platform connecting the world's luxury boutiques and brands, run by a founder whose story I admired. As a builder, I knew what great platforms look like, and this one fit the picture.
The analogy did the thinking for me. A cloud platform sells software with high margins. Farfetch kept moving into owning brands, designing and producing clothes, and carrying the economics of a retailer. In 2019 it bought New Guards, a fashion brand group, for $675 million. It never reached a sustainable profit. In December 2023, it said it could not continue without new money; Coupang took over the business, and the shareholders were left with nothing. My own losses came earlier, in 2021, but the collapse showed me exactly what the analogy had hidden.
My builder's knowledge was real. I had borrowed it for a business it did not describe.
How I build conviction now
I no longer act on borrowed conviction. That is long gone from my style. What replaced it is a simple test, and it comes down to one question.
Do I believe I can predict that this company's share of its own industry will grow over the next five years, in a market that is already growing?
Framed that way, the question forces the work. To answer it, I have to understand the competition: each rival's execution, its plans and its product lineup. Then I have to look at the company I am considering in the same way, its execution, its products and its leadership, and judge whether it is on a trajectory to gain market share or to lose it. If I do not know an industry or a company well enough to answer that question, then I do not know enough. I cannot build conviction, and I do not invest.
The second test is leadership, and it matters just as much. I have run several technology companies myself, and I know that in this industry, execution is everything. The same strategy succeeds or fails depending on the people carrying it out. If I do not like the execution, the leadership, or the CEO or founder of a company, I cannot build conviction in it, however good the story sounds.
The same discipline is where Runtime came from. A system cannot borrow conviction either. Every idea it acts on has to be written down precisely enough for a machine to test, with a defined risk budget before it is allowed to earn a return. In a sense, Runtime is the investor I became after I stopped borrowing.
Conviction has to be earned. Borrowed conviction never works.
Engineered compounding,
Deniz Erkan
Disclosure: This letter describes my personal investment history. Company examples are my own positions and are not recommendations. This history is separate from the Runtime fund and its algorithmic strategy. Past performance does not predict future results. This is not investment advice.