Everyone talks about win rate and profit factor. Nobody talks about how much you’re paying to trade — or, if you’re short, whether funding pays you back or adds to the bill. We measured both, first under a flat-rate funding model and then with real rates.
Correction (2026-08-12): An earlier version of this article presented funding as a cost for these SHORT strategies and added it to the fee bill. That read the engine’s sign convention backwards: for shorts, the funding line is negative because it is received, not paid — shorts collect funding whenever the rate is positive, which is its long-run norm. We re-ran both presets and rewrote every number below from those runs (2026-08-12, preset defaults). Same-day follow-up: the MACD column was re-run at its stated SL 12% / TP 10% after we found the first pass had silently fallen back to engine defaults (SL 8 / TP 6) — a request-field mismatch on our side, not an engine bug.
Two Strategies, Real Numbers
We ran BB Squeeze SHORT 1H (SL 10% / TP 8%) and MACD Cross SHORT 4H (SL 12% / TP 10%) on the top-50 coin universe over ~2.8 years (Nov 2023 – Aug 2026). Same cost model: 0.02% maker / 0.05% taker per side, plus a funding assumption. That funding assumption is the part that has since changed — see the re-measure notice below.
| Metric | BB Squeeze SHORT 1H | MACD Cross SHORT 4H |
|---|---|---|
| Total Trades | 2,416 | 3,511 |
| Gross Return | ~22.70% | ~54.38% |
| Fees Paid | -5.03% | -7.02% |
| Funding Received | +1.92% | +9.14% |
| Net Return | +19.59% | +56.50% |
| Fees as % of Gross | 22.2% | 12.9% |
| Net Cost Drag (fees − funding) | 3.11% | -2.12% |
BB Squeeze made ~23% gross but kept ~20% after costs. Fees alone took 5 percentage points — over a fifth of the gross profit gone to the exchange. Under the flat-rate model, funding gave back +1.92%, softening the bill.
MACD Cross 4H traded more (3,511 vs 2,416) and held positions longer (4H timeframe = more funding periods per trade). Those long holds collected +9.14% in funding — more than its entire 7.02% fee bill. Under that flat-rate model its net cost drag came out negative. That «funding paid the whole fee bill» reading does not hold in the re-measure below: with real rates this short still received funding, but less than its fees.
Funding: Income for These Shorts, but Not Enough to Cover Fees
Fees are always a cost. Funding depends on which side you’re on.
| Line Item | BB Squeeze 1H | MACD Cross 4H |
|---|---|---|
| Trading Fees | -5.03% | -7.02% |
| Funding | +1.92% | +9.14% |
| Net Drag (fees − funding) | 3.11% | -2.12% |
⚠ Re-measured 2026-09-05 — still income, but not enough to cover fees
The table above was produced on 2026-08-12 with a flat 0.01%-per-8h funding model. Two days later, on 2026-08-14, the engine switched to actual historical funding rates (methodology). The article did not follow.
Re-run today against production (
POST https://api.pruviq.com/simulate, preset SL/TP, top-50, 5x, today’s roster and window):
BB Squeeze SHORT 1H MACD Cross SHORT 4H total_trades2,367 3,356 total_funding_pct−0.93% −20.26% Sign convention:
total_funding_pctis a cost — positive means paid, negative means received, and for each trade net return = gross return − fees − funding (methodology). An earlier version of this box read the negative values the wrong way round and said these shorts paid funding. They did not.Both came out negative, so with real rates both shorts still received funding: about 0.93% (BB Squeeze) and 20.26% (MACD Cross). The flat-rate model had the sign right. What does not hold is the claim below — that MACD Cross earned back its whole fee bill and then some. In this re-run its funding income (20.26%) is smaller than its fees (
total_fees_pct33.56%), and BB Squeeze’s (0.93%) is far below its fees (24.66%). That claim is retracted.The 2026-08-12 numbers stay as the record of what we measured then. They are not today’s values, and the two runs are not directly comparable: this re-run uses today’s roster and window and normalizes fees differently, so compare the sign across runs, not the magnitude.
Reproduce:
curl -s -X POST https://api.pruviq.com/simulate -H 'Content-Type: application/json' -d '{"strategy":"macd-cross","direction":"short","sl_pct":12,"tp_pct":10,"timeframe":"4H","top_n":50,"leverage":5}'
Perpetual futures charge funding every interval (typically 8 hours). When the rate is positive — the usual state, since perpetuals tend to trade at a premium — longs pay shorts. When it’s negative, shorts pay longs. Both strategies here are SHORT, so under the flat positive rate in our model, every 8 hours in a position earned funding.
MACD Cross holds through many 8-hour marks on its 4H timeframe, so its funding line is the larger one in both runs. Under the flat-rate model it showed +9.14% received over 3,511 trades and BB Squeeze on 1H +1.92%. With actual historical rates (re-measure above) the engine reports −20.26% and −0.93% on its cost axis. Negative means received: these shorts still collected funding on net, just not enough to cover their fees.
The mirror image applies to LONG strategies: the same holding pattern that pays a short bleeds a long. Direction decides the sign.
What This Means for Your Strategy
A strategy with PF 1.17 and 2,000+ trades sounds profitable. But look at the cost structure:
- Gross PF at these fee levels becomes a thinner edge — fees scale with trade count, relentlessly
- Higher frequency = more fees — 2,416 trades × two sides each is a fee event stream no win rate can ignore
- Longer holds = more funding exposure — income for shorts in a positive-rate regime, cost for longs. The 4H short had the larger funding line here in both runs, and it was received in both
The break-even point shifts. A strategy that shows +5% gross return might be negative after fees — and for a slow short, funding can move the result either way, depending on the real rate path.
How to Reduce the Damage
- Trade less frequently — Fewer trades = fewer fee events. But too few trades means low statistical significance.
- Use limit orders — 0.02% maker vs 0.05% taker. On 2,000 trades, that’s a meaningful difference.
- Know your funding side — Positive rate: longs pay shorts. Negative rate: shorts pay longs. Check the real rate history and your direction before assuming funding is income or a cost — for these shorts it was net income under both the flat-rate model and real rates, but with real rates it did not cover the fees.
- Get fee discounts — OKX cuts futures fees by 20% through its referral program. On 2,000+ trades, that compounds. See the fee comparison at PRUVIQ.
The Number to Remember
For high-frequency strategies on crypto futures: expect roughly 5-8% of your capital to go to fees over a multi-year period (these two runs paid 5.03% and 7.02%). If your strategy doesn’t gross more than that, you’re trading for the exchange, not for yourself — and funding can add to that bill, for shorts too.
Run your own cost analysis on the PRUVIQ Simulator — every result includes fees and funding breakdowns.
Affiliate disclosure: PRUVIQ earns a commission from exchanges — currently OKX — when you sign up through our referral link. This does not affect your fee rate. Not financial advice.