Skip to content
All library documents

Calculating the Gain Needed to Recover Trading Losses

Article Quant Q&A · Author: dragonmnl

Summary

The document explains why equal percentage losses and gains do not cancel when returns compound. After a loss, the same percentage gain applies to a smaller capital base, so the account remains below its starting value. It gives a formula for the number of equal-sized gains needed to offset repeated equal-sized losses, using logarithms to solve for the required count.

It also derives the single-period recovery return after one loss: the required gain is the loss fraction divided by the remaining capital fraction. The examples are algebraic rather than empirical, and assume returns compound multiplicatively with no fees, deposits, withdrawals, or other changes in capital. These formulas describe break-even recovery only; they do not estimate how likely recovery is or how long it might take.

Key ideas

  • Equal percentage gains and losses do not offset because each applies to a different capital base.
  • The recovery gain after a loss is calculated relative to the reduced account value.
  • For repeated fixed-size losses and gains, compounding can be expressed as a product of growth factors.
  • Logarithms solve for how many fixed-size gains are needed to offset a given number of losses.

Tags

Full text
# How to calculate necessary gain to compensate a loss in a financial transaction?


# How to calculate necessary gain to compensate a loss in a financial transaction?












(Feel free to suggest the correct Stackexchange community - or otherwise - if this is not the correct one)

When trading financial markets, a gain of `x`%, won't recover a loss of `x`% (same applies to any gain-loss scenario, in any market, I assume)

e.g. I start with 100, I lose 2% (98% or original capital), I gain 2%. I'm still in loss (99.96% of original capital)

Is there a formula to calculate the necessary % gain of a single transaction over `k` loss of x% gain after y losses of x%? (if the formula is generalized to `z` gains of x% and `k` losses of x% is preferable - x% is fixed)

## Answer by Magic is in the chain (score 5, accepted)

https://quant.stackexchange.com/a/46105

Let x represent the percent change-e.g. 2%, let k represent the number of decreases, and z the number of increases. Something like this? We want to find z such that:

$\left(1-x\right)^k\left(1+x\right)^z=1$

Rearrange,

$\left(1+x\right)^z=\frac{1}{\left(1-x\right)^k}$

And take log:

$z \ln \left(1+x\right)=-k \ln \left(1-x\right)$

and solve for z:

$z =-k \frac{\ln \left(1-x\right)}{\ln \left(1+x\right)}$

Or you want z to represent the percent increase such that:

$\left(1-x\right)^k\left(1+z\right)^k=1$

$z=\frac{1}{1-x}-1=\frac{x}{1-x}$

## Answer by David Addison (score 3)

https://quant.stackexchange.com/a/46118

This is a common question type on the GMAT.

Suppose you lose x% of your initial funds, $V_0$, so you now have $V_0(1-x/100)$. What y% return on the new funds do you need to return to your initial funds?

Stated this way, we express our problem as:

$V_0(1-x/100)(1+y/100)=V_0 $, or $(1-x/100)(1+y/100)=1$

Solving for y, we have:

$y= \frac{100}{1-x/100}-100$, or

$y=\frac{x/100}{1-x/100}$

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.