Initial Capital

What is Initial Capital?

Initial Capital is the dollar amount the simulated portfolio begins with on the first day of the backtest window. From that opening balance, the simulation buys positions according to the screen and the active weighting rule, holds them through the period, and rebalances on the schedule the user has chosen. At the end, the portfolio has a closing value, and that value is reported in dollars alongside every percentage figure on the results page.

The setting is one of the simplest knobs in a backtest. Doubling Initial Capital from one hundred thousand to two hundred thousand will roughly double the closing dollar balance, but the cumulative return, the annual return, the Sharpe ratio, the volatility, the max drawdown, and every other percentage and ratio metric will be unchanged. The math is plain: percentage returns are scale-invariant, and Initial Capital is a scale.

That said, Initial Capital is not as inert as the math suggests. It governs the dollar value of each position, and dollar values interact with transaction costs and share rounding in ways that pure percentage thinking misses. A portfolio that holds twenty positions at a one hundred thousand dollar starting balance is buying five thousand dollars of each name. At five hundred dollars per share, that is ten shares; at five thousand per share, it is one share with no headroom. Those rounding effects are real, and they are visible only in dollar-level simulations.

Why Initial Capital Matters in Backtesting

Initial Capital matters in three concrete ways.

First, it drives the realism of execution. A backtest that buys one share of an expensive stock and a thousand shares of a cheap one is making decisions on integer share counts. At very small starting balances, the simulation may not even be able to take a meaningful position in stocks above a certain price, because a single share would breach the position-size cap or distort the intended weighting. Bumping Initial Capital up gives the simulation room to express the weighting rule cleanly, so the resulting allocation more faithfully reflects what the strategy is supposed to do.

Second, Initial Capital interacts with any fixed-cost trade assumptions. If a backtest applies a flat dollar amount per trade, that flat cost is a much heavier drag at one thousand dollars of starting capital than at one hundred thousand. Modern brokerage accounts mostly use either zero commissions or a percentage of trade value, which makes this concern less acute than it once was, but it still matters whenever a study includes a fixed minimum or any per-share charge.

Third, Initial Capital sets the scale of every reported dollar figure. The closing balance, the cumulative profit chart, the dollar-denominated drawdown, the annual gain or loss in dollars, and any per-period dollar metrics all scale linearly with the starting balance. For users who think in dollars rather than percentages, this changes how the result feels even when the underlying performance numbers are identical.

How SledgeKey Implements Initial Capital

The configuration panel includes an Initial Capital field that accepts a dollar value. The simulation opens its first rebalance with that balance on the start of the chosen window, allocates across positions according to the active weighting rule and any caps the user has set, and reports both percentage and dollar figures in the results.

Changing Initial Capital does not affect the screen, the universe, the rebalance schedule, or the underlying point-in-time data. It changes only the scale at which the simulation operates. Two backtests with the same screen, same window, same rebalance frequency, same weighting metric, and same transaction cost assumption will produce essentially identical percentage returns regardless of what value is entered for Initial Capital, with the small caveats around share rounding noted above.

For most analyses, the default is a reasonable choice and there is no benefit to changing it. Initial Capital becomes worth attention in two cases: when modeling a strategy a user actually intends to fund at a known dollar size, and when studying how a strategy behaves at very small or very large dollar scales, where execution frictions become more visible.

Common Pitfalls

A frequent mistake is treating Initial Capital as a lever to improve a backtest. Because the headline metrics are scale-invariant, raising or lowering the starting balance to chase better numbers achieves nothing. The Sharpe ratio of a strategy is the same at ten thousand dollars and at ten million; only the absolute dollar figures change.

A subtler pitfall is assuming results scale perfectly. They mostly do, but not exactly, because of share-rounding artifacts and any fixed-cost trade assumptions in the cost model. Two backtests with identical configuration but different Initial Capital values can show small differences in cumulative return at the second or third decimal place, and those differences are real, not noise. They reflect the lumpiness of integer share allocation.

A third pitfall is reading dollar figures from a backtest as a forecast of dollar outcomes. The percentage return is the strategy's signal; the dollar figure is just that percentage applied to the chosen starting balance. A ten percent annual return on one million dollars produces one hundred thousand dollars of profit on paper, but the strategy did not produce one hundred thousand dollars. It produced ten percent, and one million was a property of the input, not the strategy.

Related Concepts