Leverage and Compound Growth Mathematics

Leverage is the most misunderstood tool in trading. Used carefully, it scales a genuine edge into meaningful returns. Used carelessly, it converts a profitable strategy into guaranteed ruin. The reason is mathematical: compound growth is geometric, and leverage interacts with variance in a way that punishes overuse.

Arithmetic versus geometric returns

If a strategy returns $+50%$ then $-50%$, the arithmetic average is $0%$, but the geometric result is:

$$(1 + 0.5)(1 - 0.5) = 0.75$$

A 25% loss of capital. The variance drag is symmetric. Geometric growth is always less than the arithmetic mean by approximately:

$$g \approx \mu - \frac{\sigma^2}{2}$$

Higher variance means more wealth destroyed by the path of returns.

How leverage interacts with the drag

Leverage $L$ multiplies both mean and variance. The leveraged geometric growth becomes:

$$g_L \approx L\mu - \frac{L^2 \sigma^2}{2}$$

The variance penalty scales with $L^2$ while the return scales with $L$. There is a leverage level that maximizes geometric growth — and beyond it, growth turns negative despite a positive underlying edge. The optimal leverage is:

$$L^* = \frac{\mu}{\sigma^2}$$

This is the continuous-time Kelly fraction for a Gaussian return process. Trading above $L^*$ guarantees negative geometric growth in the long run, even with a real edge.

A concrete example

Suppose a strategy has annualized mean return $\mu = 15%$ and volatility $\sigma = 20%$. Unleveraged geometric growth:

$$g \approx 0.15 - \frac{0.04}{2} = 0.13$$

Optimal leverage is $L^* = 0.15 / 0.04 = 3.75$. At 8x leverage:

$$g_{8x} \approx 1.20 - \frac{64 \cdot 0.04}{2} = -0.08$$

Growth turns negative despite a profitable underlying strategy. Over-leverage converts edge into ruin.

Practical implications

• Estimate $\mu$ and $\sigma$ conservatively. Optimistic $\mu$ leads to over-leverage.
• Trade well below $L^*$ because parameter estimates carry error and markets are non-stationary.
• Treat leverage as a risk multiplier, not a return multiplier. It scales drawdowns as fast as it scales gains.
• Account for funding costs. Borrowing at rate $r$ shifts growth down: $g_L \approx L(\mu - r) - L^2\sigma^2/2$.
• Cap absolute leverage regardless of what the math suggests. Survival matters more than theoretical optimality.

The durable lesson: leverage has a mathematical ceiling, and the market does not warn you when you cross it. Conservative leverage applied to a real edge beats aggressive leverage applied to the same edge every time, over any horizon long enough for the math to express itself.