Rounding Down: What Happens to Risk When the Lot Step Is Coarse

[Fairman]

Rounding Down: What Happens to Risk When the Lot Step Is Coarse

Scenario: a 3,800-dollar account, 1 percent risk, EUR/USD, stop 25 pips. Your broker's lot step is 0.01.

Work it out. Risk amount is 38 dollars. Risk per standard lot at 25 pips is 250 dollars. The exact lot size is 38 divided by 250, which is 0.152.

You cannot trade 0.152. Your options are 0.15 or 0.16. At 0.15, the risk is 0.15 times 250, which is 37.50 dollars, or 0.99 percent. At 0.16, the risk is 40 dollars, or 1.05 percent.

The disciplined choice is 0.15, because rounding down never exceeds the plan.

Now imagine a trader who rounds up every time. Across 100 trades, the average overshoot might be 3 or 4 percent of the intended risk. That is small per trade, but it never errs in your favor. Over time it makes your real risk larger than your written risk, and your statistics become slightly inconsistent.

The effect is bigger for small accounts. On a 300-dollar account, a lot step of 0.01 may be a large fraction of your intended size, and rounding errors of 20 or 30 percent are possible.

Takeaway: always round down, and record the actual risk, not the planned risk, in your journal.

Practical step: for your last ten trades, compute the planned and actual risk and see whether you tend to overshoot.

[FXS HongKong Scalper]

Good example of why rounding down is the safe default. A related trap is the minimum lot. On a small account with a wide stop, even 0.01 can exceed the planned risk, and then the only disciplined options are a tighter stop with a different setup or no trade.

Some platforms also round lot sizes automatically to the nearest step rather than down. It is worth testing with your own EA or calculator that it truncates.

[Fairman]

The point that rounding up never errs in your favour is subtle but important. A random rounding error would average out over many trades. A systematic error, always rounding up, accumulates. If the average overshoot is 3 or 4 percent, a trader who believes they risk 1 percent per trade is actually risking about 1.035 percent. Over a long losing streak, that adds up to a deeper drawdown than planned. More importantly, it means the journal's planned risk figures don't match reality, so any analysis based on them is slightly off. Rounding down keeps the plan honest.

[Amanhasnoname]

This one bit me on a 1,500 dollar account where I wanted to risk 1 percent on a 40 pip stop on GBPJPY. The calculator gave a lot size of about 0.0096, so the broker minimum of 0.01 looked like a harmless rounding. In reality that was already above my plan, and with the stop widened a little during volatile hours I was risking nearly twice what I had written in my journal.

What fixed it was adding a column called actual risk next to planned risk and filling it from the order ticket, not from my calculator. After a month I could see that my average overshoot was about 12 percent, mostly on the small trades where the lot step is a big slice of the size.

My rule now is that if the minimum lot already breaks my risk limit, I either use a tighter stop on a better level or I skip the trade. Skipping feels bad at first, but it keeps the statistics honest.

[Fairman]

The small account effect deserves attention. On a 300 dollar account at 1 percent risk with a 25 pip stop, the calculated size is 0.012 lots. Rounding down gives 0.01, risking 2.50 dollars, about 0.83 percent. Rounding up gives 0.02, risking 5 dollars, about 1.67 percent. That's a 67 percent overshoot, far more than a rounding error. On small accounts, the choice between rounding up and down can change the risk dramatically. Rounding down is still the right default, but it means the actual risk can be well below the plan, which affects expected returns.

[Fairman]

Recording the actual risk rather than the planned risk in the journal is the right practice. It means R calculations are based on what was really at stake. If the planned risk was 38 dollars but the actual was 37.50, a loss of 37.50 is exactly minus 1R. Using the planned figure would show it as minus 0.987R, a small difference but one that adds noise. More importantly, if stops are widened or slippage occurs, the actual loss relative to the actual risk shows the true R, which reveals execution problems clearly.

[Fairman]

The practical step of comparing planned and actual risk for the last ten trades often reveals patterns. Some traders find they consistently round up, others that they occasionally use the wrong pip value or account balance. A spreadsheet column showing actual risk as a percentage of equity, next to the planned percentage, makes the comparison automatic. Has anyone here found a consistent overshoot in their actual risk? I'd be curious how large it was, what caused it, and whether fixing it changed your results in any noticeable way over time.