How Fees And Financing Eat Your P&L

Every strategy back-test starts clean: enter here, exit there, mark the difference. Real trading has a second ledger running underneath that back-test, and it never stops: commissions on every fill, interest on borrowed cash, a fee to borrow the stock if you're short, and a spread on top of the reference rate if you're financing anything with leverage. None of these show up as a dramatic loss on any single trade. Over a year, on a book that turns over its capital many times, they can be the difference between a strategy that works and one that doesn't.

The four drags

Commissions. Charged per share or per trade, and they scale with turnover — a strategy that trades in and out ten times more often pays roughly ten times the commission drag for the same market exposure.

Margin interest. If you're using leverage, you pay interest on the borrowed portion, typically a reference rate plus a spread the broker sets based on your relationship and collateral.

Borrow fees (stock loan). To be short a stock, someone has to lend it to you, and that's not free. Easy-to-borrow names cost a few basis points a year; hard-to-borrow names — heavily shorted small-caps, for instance — can cost 10, 20, even 50+ percent annualized.

Financing spread. Any leveraged or derivative position embeds a financing rate (think the implied rate in a future's basis, or the funding leg of a swap), and that rate is set by the dealer with a spread built in, not the "clean" risk-free rate a textbook model assumes.

Worked example

You run a long-short pair: long $1m of an easy-to-borrow large-cap, short $1m of a stock that happens to be hard to borrow at 8 percent annualized. The trade is on for 40 trading days, roughly two months.

• Long side financing (leverage isn't used here, ignore).
• Short side borrow: $1{,}000{,}000 \times 0.08 \times (40/252) \approx 12{,}698$, about $12,700.
• Commissions: 200,000 shares round-tripped across both legs at $0.005/share $= 1{,}000$, about $1,000.

Total drag: roughly $13,700 on a $2m gross position held two months. If the pair's raw price convergence made you $18,000, your actual P&L after costs is closer to $4,300 — a 76 percent haircut. A back-test that ignored the borrow fee on the hard-to-borrow leg would have shown a much better Sharpe than the strategy can actually deliver.

What to do about it

Get borrow rates checked before you put on a short, not after — a name that's easy to borrow today can go hard to borrow the day a short squeeze starts, and the fee resets to whatever the lender wants to charge. Track financing drag as its own line in your P&L attribution rather than burying it in "other," because a strategy that looks fine gross but bleeds on financing is telling you either to trade less often or to avoid the specific hard-to-borrow names.