# Mistake in my stastistics?

Mathew I was just wondering on my new signal Conservative Growth. It is showing a 41% annulized return but a negative profit per unit? Since the signal is making money shouldn’t that be a positive figure? Thanks

The P/L per unit is first computed per trade, and the outcomes of this are averaged across trades. This can be negative even if the net profit is positive. I think it would be better to compute the P/L per unit as the net profit divided by the total number of units traded. The outcomes of these two formulas will be identical if the number of units is uncorrelated with the P/L per unit or when the number of units is constant across trades, but usually not in other cases.

OK thanks Jules. Still clear as mud. However I should have thought about things like this win I left school.

I was brief because it has been discussed several times on the forum in the last year. I should have said that. You can find more detailed explanations there.

A simple example:

trade 1: 100 units with \$1 profit per unit

trade 2: 1 unit with \$2 loss per unit.

Then the total profit is 100 * \$1 + 1 * (-\$2) = \$98. The total number of units is 100 + 1 = 101. So you and I would think that the P/L per unit is \$98 / 101. But C2 computes it as the average of \$1 and -\$2, which is negative (-\$0.50). So it does not take into account that the position size of the first trade was much larger.

Even if a subscriber trades only 1 unit per trade, he would not necessarily get the C result if the trade consists of multiple legs, because the subscriber would take only the first leg. See the previous discussions.

Jules I appreciate the time you spend explaining this to me. Thanks