A farm goes looking for working capital. The rate comes back somewhere between unpleasant and punishing, or the money does not come at all and the cycle gets financed the way it usually does — against next harvest, from someone in the chain who will take their cut in the price instead of in interest.
The usual explanation is that lenders do not understand aquaculture. I used to believe that. I no longer think it is the main thing going on.
Lenders price uncertainty, not badness
Put yourself on the other side of the desk. You are asked to advance money against a biological asset you cannot inspect, on land you have not visited, managed by someone whose track record exists mostly in their own memory. You are not being asked to judge whether this is a good farm. You are being asked to judge something you cannot see at all.
So you price the uncertainty. Not maliciously — arithmetically. When the range of plausible outcomes runs from excellent to total loss and nothing in the file narrows it, the only rational move is to price near the bad end and take the deal or leave it.
Which means the well-run farm and the shambolic one are quoted roughly the same. An unreadable farm is not a bad risk. It is an unassessable one, and unassessable is expensive.
The data exists — it just is not legible
Here is what makes this genuinely maddening. The information a lender would need already exists. Every farm I have ever walked knows its own numbers. Stocking densities, feed conversion, survival by pond, which manager runs the tightest cycle, what happened in the bad year and why.
It lives in a notebook, a spreadsheet somebody rebuilds each season, a WhatsApp thread, and the head of a unit manager who has been there eleven years. All of it real. None of it legible to anyone standing outside the gate — and crucially, none of it in a form that survives the manager leaving.
So the farm carries the full cost of its own good practice and captures none of the credit for it. That is the same structure as a hatchery that cannot prove its batch at the gate and a producer who cannot collect the premium. Three different rooms, one missing floor.
Record-keeping is a balance-sheet item
The reframe I would push on any operator: stop filing bookkeeping under admin. Under these conditions it is the thing that converts your actual performance into something an outsider can price. That is not overhead. That is the asset.
What it takes is a production record built as the work happens rather than reconstructed when somebody asks — events captured in a fixed shape from the first entry, attributable to the person who recorded them, and under the farm’s own control as to what gets shared and with whom. That last part matters more than it sounds. A record you cannot withhold is not an asset, it is an exposure.
I am not going to tell you this makes credit cheap. Rates are set by people with their own constraints, and a legible farm having a better conversation is a mechanism, not a promise. But the farm that can show three years of its own operating history is having a different conversation from the one that cannot — and right now almost every farm is having the second one.