Ten years of trading, and I lost most of the money to myself

Opinion and personal account. This is my story, not a manual. The figures from studies are sourced and linked; my own story is a single case and proves nothing. That is exactly why I am telling it.

The balance sheet I do not enjoy writing down

I have traded continuously for about ten years, since 2015/16, crypto and equities. On balance I have won more than I lost.

And so that this does not sound like skill, here is the concrete reason: I bought Bitcoin at 500 euros and held it. The only decision that genuinely paid off was one where I did nothing. No timing, no analysis, no discipline — patience and luck, in that order.

And now the part that almost never comes in this industry: I cannot prove it to you. There is no screenshot here, and that is deliberate — a screenshot is the most worthless currency in this trade, anyone can make one, and whoever shows you one wants to sell you something. So take the figure for what it is: the unevidenced self-report of a man who is about to explain why self-reports are worth nothing. It carries no argument in this text. Everything that matters here is below, with a DOI, and is verifiable without me.

Everything I did actively and with conviction over those same ten years did not improve that result. It ate into it.

And yet: most of the money I lost over those ten years, I did not lose to bad analysis.

I lost it to myself.

That is not modesty as a pose. It is the most accurate description I have for the process. The analysis was often right. The tips I received in my mid-twenties in Berlin from friends in the actual financial world — people moving millions — turned out right almost every time. I lost anyway.

I sold too early because I was afraid. I held too long because I hoped.

The mistake has a name, and it is older than I am

The first time I told someone this, I got an unwelcome piece of information: my personal drama is a textbook chapter.

Shefrin and Statman described the pattern in 1985 and gave it a name: the disposition to sell winners too early and ride losers too long. They do not attribute it to stupidity but to a bundle of very human mechanisms — mental accounting, regret aversion, weak self-control — and they show that tax motives alone cannot explain the observed patterns.

Thirteen years later Terrance Odean measured instead of theorising. He analysed the trading records of 10,000 accounts at a large discount brokerage. The result: those investors showed a strong preference for realising winners over losers. It was not portfolio rebalancing. It was not the trading costs of low-priced stocks. And — the sentence that stings — it was not justified by subsequent performance. For taxable accounts the behaviour was simply suboptimal.

So my mistake was not original. It was the most common one there is.

What 66,465 households reveal about me

The second study I should have read earlier is called exactly what it is: “Trading Is Hazardous to Your Wealth”. Barber and Odean examined 66,465 households at a large discount broker between 1991 and 1996.

The households that traded most earned 11.4 percent annually. The market delivered 17.9 percent over the same period. The average household reached 16.4 percent while turning over 75 percent of its portfolio per year.

The gap did not open because these people picked the wrong stocks. It opened because they traded. The explanation the authors offer is overconfidence: those who believe they know more, trade more — and pay for it.

I have read that paragraph several times and recognised myself in it every time.

Why the knowledge did not help

And now the part I actually care about.

I knew all of this. Not every study by name, but the mechanics. I could explain to you why people let losses run and cut gains short — the cause lies in prospect theory by Kahneman and Tversky: a loss weighs more heavily than an equivalent gain feels good. As long as the position is open, the loss is only a number on a screen. Sell, and it becomes true. So you do not sell.

I understood that intellectually. The execution: rather mediocre.

That is the whole joke, and it took a decade to land. Knowledge and behaviour are two different systems. One works at the desk, calmly, with coffee. The other works at 11 p.m., when the position is red, your pulse is up, and hope makes a better argument than any study.

My friends from back then grew richer than I did. Not because they knew more — because they had themselves under control.

The basics I skipped

When I go through my most expensive moments, there is never a complicated misjudgement. There are skipped obvious things:

1. No stop-loss set. Not because I did not know what one is. Because in that moment I believed I had spotted the exception.
2. No look at the fear index. By which I mean a sentiment gauge such as the crypto Fear and Greed Index, which compresses market sentiment onto a scale from 0 (extreme fear) to 100 (extreme greed). A glance at it would have told me, on more than one purchase, that I was doing precisely what everyone else was doing. Whether such indices have any forecasting power is a separate question I explicitly do not answer here — and one we test open-endedly in the lab. As a mirror for my own state, it would have been enough.
3. Position size by feel. A little more when I was convinced. Exactly when conviction is worth least.
4. The hundred small disciplines that keep you from the crash and all share one property: at the decisive moment they are annoying.

None of this is secret knowledge. All of it is in every beginner’s book. Which is what makes it bitter.

What the studies do not support

I quoted numbers above, so the counter-argument belongs here too — otherwise I make the very mistake I accuse others of.

The most obvious objection first: these studies come from US discount-brokerage data of the nineties. No crypto, no smartphone app, no 24/7 markets, no zero commission. So one might hope the problem has since taken care of itself.

That hope has been examined. In 2022 Barber, Huang, Odean and Schwarz analysed the users of a commission-free trading app. Those investors engage in more attention-induced trading than other retail investors — partly because the app attracts less experienced people, partly because of its particular design. And intense buying by those users forecasts negative returns: for the most-bought stocks of a given day, the average abnormal return over 20 days was −4.7 percent.

So the app did not fix the effect. It accelerated it.

Even so, and this is the honest limitation: the 11.4 versus 17.9 percent is a cross-sectional average across tens of thousands of households — it does not prove that trading is the cause of your personal return. Overconfidence is a plausible explanation, not the only one. The Robinhood analysis measures attention effects, not my disposition effect; it supports my story thematically, not evidentially. And while the disposition effect is among the better-replicated findings in behavioural economics, “better replicated” in this field does not mean “settled”.

And my own story? A single case. It is the occasion for this text, not its evidence. If you derive a rule for your portfolio from my ten years, you have read it wrong.

What I do differently today — and what you get from it without my system

The most useful part of this text costs nothing and needs no software. These are the consequences I drew from the four points above, and they work on a slip of paper exactly as well as in code:

1. The stop exists before the position does. Not afterwards, not “if things go badly”. Beforehand, in writing, with a number. After you enter, your judgement is compromised — you now hold an opinion and an interest.
2. Position size is a calculation, not a feeling. The only sensible question is: what may this single trade cost me if I am wrong? Size comes after that answer, not before.
3. “This time it’s different” is a symptom, not an argument. When that sentence shows up, you are not smarter than usual — you are in the middle of the disposition effect. For me it was the most reliable early indicator of an expensive day.
4. Calculate the friction before you enter. Cost per round times number of rounds. That figure decides the outcome more often than direction does.
5. Write the rule down while you are calm. The only person who can overrule you at 11 p.m. is the person you were at 11 a.m. Give him something in writing.

Point 5 is the whole trick, and it is why this turned into software for me: what you automate does not negotiate with you. A note beside the screen is already automation — it is merely easier to ignore than a program.

The consequence: rules that hold when I do not

Exactly one thing follows from this balance sheet, and it is unromantic.

If knowledge was never the problem, more knowledge does not solve it. Another newsletter of tips would have done nothing for me — I had the best tips available, straight from people who lived off them. What I lacked was something that does not negotiate in the moment of emotion.

So I started building it: rules as code. A system that sets the stop because it belongs there, not because I happen to feel disciplined. First for myself, then for my family, now for friends. How and why that came about is in my story — which also explains why I am fundamentally mistrustful of this trade.

Two things belong here so this does not turn into an advertisement.

First: a system only closes the execution gap, not the physics. Discipline does not repair a broken cost model. Perfect obedience to a rule that is priced wrong just produces losses on schedule instead of at random.

I know this because our own books say so. When we audited what our trading actually cost us, the fees alone consumed more than the entire gross profit on two of the paths we examined — the arithmetic was lost before any psychology entered the room. Those figures, and the reasons they are weaker than they look, are laid out in the fee truth. The lesson for this text is narrow: everything I described above happens on top of your cost structure, never instead of it.

Second: the human still sits at the table. A system takes the execution off your hands, not the enduring. Why my father’s verdict on this entire industry was almost correct is its own chapter.

Nothing is sold here before a public track record exists. Until then you get what I have: the mistakes, the numbers, and the willingness to write both down.

More from this desk in Wealth & Mindset.