Why I'm writing this down

There is no track record on this site. That is deliberate, and this letter explains what I intend to do about it over the next two years.

Every trading website you have ever landed on opens with a number. Someone made 340% and would like to explain how. The explanation always sounds coherent, because explanations constructed after the fact always do. You are reading a story that was written backwards from a result, and there is no way for you to tell how much of it was skill and how much was a sector that happened to move.

I want to do it the other way round. The plan first, in public, with a date on it. Then two years of results verified by someone with no reason to make me look good. If it works, you will have watched it happen rather than been told about it afterwards. If it doesn't, that will be here too, in the same table, in the month it happened.

What this actually is

One account. My own money. No fund, no clients, nothing for sale, and — for the avoidance of any doubt — no course at the end of it.

I buy equities in capital-intensive cyclical industries: crude and product tankers, dry bulk, containers, LNG carriers, offshore drilling, oilfield services, refining, gas, coal, uranium, copper. I go both long and short. The book is concentrated and it uses leverage. Those last two facts will do a lot of work later in this letter.

From January, the account is entered in the US Investing Championship in the million-dollar-plus division. Brokerage statements go to the organisers every month; they publish the standings; I don't. That's the entire appeal of it. It's a verification mechanism I don't control.

Why these sectors and not better-behaved ones

Because almost nobody is looking at them, and because the assets underneath are things you can count.

A ship has a replacement cost. So does a rig, so does an orebody. You can find out what it costs to build one, how long the yard queue is, how old the existing fleet is, and how many are being scrapped. None of this is secret — it is published weekly and read by almost no one, because these industries have been so bad for so long that the analysts covering them have been reassigned to something with better meetings.

That is the whole opportunity. Not cleverness. Attention. When capital has been starved out of an industry for a decade, nobody orders new supply, and when demand eventually turns up against a fleet that nobody replaced, the numbers get absurd. Then the analysts come back, and by that point it's over.

The screen never says "this is cheap because the cycle turned." It says the same thing at the top as it does at the bottom.

Which is why a low multiple in a cyclical is usually a warning rather than an invitation — it means the market is looking at peak earnings and correctly refusing to believe them. Getting that distinction right is most of the job, and I'll be writing about it a lot.

The expensive thing I've learned

Being wrong is survivable. Being right too early is what actually costs money.

You can build the thesis correctly, identify the supply constraint, size the position, and then sit through two years of the market not caring — during which the position can fall by half without a single fact in your argument turning out to be false. The difference between being right and being paid for it is measured in years, and everything about how I now run risk comes from having lived in that gap.

It is also the reason I am wary of my own conviction. The strongest feeling I have ever had about a position has not reliably been the most profitable one. It has, however, reliably been the largest one, which is a different and more dangerous fact.

What you'll get, roughly monthly

• What I bought and sold, and the reasoning as it stood at the time
• What moved against me, and whether the thesis or the timing was wrong
• What I'm reading in the freight and rig data
• From January, the verified figure for the month, whatever it is

No alerts. No price targets. No suggestion that you should buy or sell anything — I have no idea what your situation is, and anyone writing a public newsletter who claims otherwise is selling something.

The part I'd rather leave out

This approach produces deep drawdowns. Concentrated positions in leveraged cyclical equities go down 30% or more in weeks, sometimes with nothing at all wrong with the underlying case. There will be months in that table that look bad, and at least one stretch that looks alarming.

I'm saying so now, in the first letter, because the alternative is to say it later when it's happening and it sounds like an excuse. A manager who presents this kind of strategy as smooth is telling you something about the manager rather than the strategy.

What happens next

Between now and December there will be a few more letters, all about method rather than results — how I read the supply side of these industries, what I think the market consistently gets wrong about them, and the mistakes that taught me the most.

Then January, and the clock starts. If you want to watch a record being built from the first month rather than reading about one afterwards, that's what the list is for.