What the copper curve taught me about patience

Six weeks ago I didn't know what a warrant was. I still couldn't give you a clean definition. But the copper curve has been explaining something to me since June, and the thing it's explaining is mostly about time, not metal.

Everything hangs on one fact: a financial asset can just be a number in a ledger. A tonne of copper concentrate can't. It has to actually be somewhere a warehouse, a ship, a queue outside a smelter and every day it sits there, someone is paying for it to sit there.

That cost is where the curve comes from. When the forward price is above spot (contango), the market is basically paying you to wait, because holding the metal costs money and somebody has to cover it. When forward is below spot (backwardation), the market wants the metal now and will pay up to jump the queue.

I spent my first two weeks reading the curve as a forecast, like it was telling me where copper was going. It isn't. It tells you what waiting costs right now, which is a completely different thing, and I mixed the two up for longer than I'd like to admit.

What operations actually is

My job sits between the traders and the books that keep the traders honest. Mark-to-market and hedging reports every day. Reconciling the physical positions against the financial ones. Long/short maintenance on client accounts. Some research on junior mining projects that might feed the pipeline one day.

I came in picturing a desk as a place where people make calls. From where I sit it's mostly a place where people reconcile, and I've gotten weirdly into the reconciling, because that's where you find out what's actually in the book.

The two trades

A few weeks in I was running an off-schedule reconciliation and found two duplicate executions. 800 tonnes booked twice, sitting somewhere it should only have been once. Flagging it took a five-minute conversation.

What stuck with me wasn't the catch. It was that nothing had gone wrong yet. Nobody had lost money. It was just a position that was going to be wrong later, on some day when being surprised by it would have cost a lot more than it did that Tuesday morning.

Most of the errors worth catching haven't cost anything yet. Almost all the checking I've been handed has that shape. You're not cleaning up damage, you're removing the conditions for it before it shows up. Deeply unglamorous. I'm starting to think it's most of what a good book is.

On patience

This is the part four years of school prepared me badly for. Market opens and my instinct is to have a view immediately, ideally an interesting one.

The desk doesn't move like that. A cargo takes weeks. An assay takes however long an assay takes. Twice now I've watched the price confirm something a trader said a week earlier, and both times the thing that got me wasn't that they were right. It was that they'd sat with the view for a whole week without needing the market to agree with them in the meantime.

I can't do that yet. I still want to be proven right by end of day. But I've at least started noticing that I want it, and apparently the noticing is step one.

Six weeks in, four things I'm keeping

• Ask what a number is for before you ask whether it's right. Most of the reports I touch have a decision behind them, and the decision sets how precise you actually need to be.

• Friction in physical markets isn't noise around the price. Freight, warehousing, assay timing some of it is the price.

• The person who reconciles the book is the one who ends up understanding the book.

• Say "I don't know" fast. At 9am it costs nothing. At 4pm it can cost a lot.

Ask me again in six months and I'd bet most of this reads as naive, which is sort of why I wanted it written down now. The curve won't care either way. It'll still be sitting there pricing the cost of waiting, not bothered in the slightest by anyone needing to be right this week.