Gross Alpha Is Not Deployable Alpha: Trading Costs, Turnover, and Market Impact

Gross alpha measures the theoretical return generated by a strategy before any trading costs are deducted. It is the number that appears in research papers, backtests, and marketing materials because it represents the pure signal quality uncontaminated by implementation friction. Net alpha, also called deployable alpha, is what remains after all costs are subtracted: commissions, bid-ask spreads, market impact, financing charges, borrow fees, and operational overhead. The distance between gross and net alpha is where the majority of apparently attractive strategies fail to make the transition from research to live deployment.

The magnitude of this gap varies enormously across strategies. A low-turnover equity strategy that rebalances quarterly and trades liquid large-cap stocks might retain eighty to ninety percent of its gross alpha after costs. A high-frequency crypto strategy that trades dozens of times per day might retain less than twenty percent. The difference is not a matter of degree; it is a matter of economic viability. A strategy that loses seventy percent of its edge to costs is not a slightly worse version of the same strategy; it is a fundamentally different proposition that may no longer justify the risk.

Investors and researchers must internalize a simple but counterintuitive fact: gross alpha is not spendable. It cannot be used to pay fees, meet liabilities, or compound into future wealth. Only net alpha serves these functions. Any evaluation that treats gross alpha as the decision metric is effectively ignoring the most important question in strategy deployment: what do I actually get to keep?

The cost of trading is not a single number but a layered system of frictions that interact in complex ways. The most visible layer is explicit costs: exchange fees, broker commissions, and regulatory charges. These are easy to measure and typically small for institutional accounts, though they can be substantial for retail traders. A typical equity trade might incur five to twenty basis points in explicit costs, while a crypto perpetual trade on a major exchange might incur two to ten basis points per side.

The second layer is implicit costs, dominated by the bid-ask spread and market impact. The spread is the difference between the best bid and best ask prices, and it represents the immediate cost of entering and exiting a position. In liquid markets like major equity indices, spreads are tight, often less than one basis point. In less liquid markets like small-cap stocks or altcoin pairs, spreads can be ten to fifty basis points or more. Market impact is the additional price movement caused by the trade itself: a large market order pushes the price against the trader, and this impact can exceed the spread for sizable positions.

The third layer includes financing and opportunity costs. Leveraged strategies pay financing charges on borrowed capital, which accumulate at rates that vary with central bank policy and market conditions. Short-selling strategies pay borrow fees that can spike during periods of high demand. Opportunity costs arise from delayed execution, partial fills, and missed trades when liquidity is insufficient. Together, these three layers can consume a substantial portion of gross alpha, and any strategy evaluation that ignores any layer is incomplete.

Turnover is the silent killer of alpha. Every trade incurs costs, and these costs compound with each round trip. A strategy that turns over its portfolio once per year faces a very different cost environment from one that turns over daily. The relationship between turnover and cost erosion is approximately linear for low to moderate turnover levels, but becomes superlinear at high turnover because market impact grows with trade size and frequency.

Consider a strategy with a gross alpha of ten percent annually and a turnover of ten times per year. If each round trip costs fifty basis points in total friction, the annual cost drag is five percent, leaving a net alpha of five percent. If turnover increases to fifty times per year, the cost drag rises to twenty-five percent, wiping out the entire edge and more. This arithmetic explains why many high-frequency strategies that look spectacular on paper produce disappointing real-world results: the costs simply overwhelm the signal.

The turnover problem is particularly acute in less liquid markets. In crypto, where liquidity is fragmented across multiple exchanges and many altcoin pairs have shallow order books, a strategy that trades frequently can easily see its execution costs exceed its theoretical edge. Even in equities, where liquidity is generally deeper, high-turnover strategies face capacity constraints that limit the amount of capital that can be deployed before market impact becomes prohibitive. The practical lesson is that turnover is not merely an operational detail; it is a fundamental determinant of whether a strategy is economically viable.

Every strategy has a capacity limit: the amount of capital it can manage before its own trading begins to erode the edge it is trying to capture. For strategies that trade liquid assets with low turnover, capacity limits may be hundreds of millions or even billions of dollars. For strategies that trade illiquid assets or require frequent rebalancing, capacity limits may be in the tens of millions. Understanding where this limit lies is essential for any allocator considering a strategy.

Capacity constraints manifest through market impact. As the strategy's order size increases relative to the average daily volume of the assets it trades, the price moves against it. A strategy that can enter a position with negligible impact at ten million dollars may see its execution price deteriorate by twenty to fifty basis points at one hundred million dollars. This deterioration is not a one-time cost; it compounds with each trade and progressively reduces the strategy's net alpha.

The relationship between size and performance is often non-linear. Below the capacity threshold, the strategy performs as expected. Near the threshold, performance begins to degrade gradually. Above the threshold, degradation accelerates rapidly, and the strategy may become unprofitable despite continued gross alpha generation. This is why some strategies work beautifully at small scale and become disasters at large scale. The alpha was never scalable; it was merely a small-capacity phenomenon that happened to work well for a while.

Implementation shortfall, first formalized by Perold in 1988, measures the difference between the return of a theoretical paper portfolio and the return of the actual portfolio after execution. It is the most comprehensive metric for capturing the gap between research and reality because it includes all sources of execution friction: delays, partial fills, price movements between decision and execution, and market impact. A strategy with a gross alpha of fifteen percent and an implementation shortfall of eight percent has a deployable alpha of seven percent, and this seven percent is the only number that matters for investment decisions.

The components of implementation shortfall can be decomposed to diagnose specific problems. Delay costs measure the price movement between the decision time and the order submission. Execution costs measure the difference between the arrival price and the average execution price. Opportunity costs measure the profit foregone on portions of the order that were not filled. Each component reveals a different aspect of the execution process, and understanding which component dominates is essential for improving deployable alpha.

For systematic strategies that generate many small trades, implementation shortfall is particularly important because the costs accumulate across hundreds or thousands of transactions. A strategy that trades fifty times per month with an average implementation shortfall of ten basis points per trade loses five percent annually to execution friction alone. This is why professional quant firms invest heavily in execution infrastructure: better algorithms, direct market access, and co-located servers can reduce implementation shortfall by several basis points per trade, which translates into millions of dollars in additional alpha at scale.

Market impact is the price change caused by a trade itself, distinct from the price change that would have occurred regardless of the trade. For small orders in liquid markets, impact is negligible. For large orders, or for orders in thin markets, impact can be the dominant cost. The relationship between order size and impact is approximately square-root: doubling the order size increases impact by roughly forty percent, not one hundred percent. But even this sublinear scaling means that large trades face materially worse execution than small trades.

Impact has both temporary and permanent components. Temporary impact is the price deviation that occurs during and immediately after the trade, which tends to dissipate as market participants absorb the order flow. Permanent impact is the lasting price change that remains after the temporary effect has faded. In liquid markets, temporary impact dominates. In illiquid markets, permanent impact can be substantial. A strategy that systematically trades in the same direction as its signals will accumulate permanent impact over time, gradually eroding its own edge.

Mitigating market impact requires sophisticated execution techniques. TWAP (Time-Weighted Average Price) and VWAP (Volume-Weighted Average Price) algorithms break large orders into smaller pieces and execute them over time to minimize market disruption. Iceberg orders hide the true size of the order by displaying only a small portion at a time. Smart order routing directs trades to the venues with the best liquidity at each moment. These techniques can reduce impact by fifty percent or more compared to naive execution, but they add complexity and require specialized infrastructure.

Leverage is a common tool for amplifying strategy returns, but it comes with costs that must be explicitly modeled. The financing cost of leverage depends on the interest rate environment and the specific terms of the borrowing arrangement. In a low-rate environment, financing costs may be negligible, adding only fifty to one hundred basis points annually. In a high-rate environment, financing costs can exceed three to five percent annually, creating a substantial drag on returns.

Short-selling strategies face an additional cost: borrow fees. When a strategy shorts a security, it must borrow that security from a lender and pay a fee for the privilege. Borrow fees vary widely depending on the security's availability and demand. Large-cap stocks in liquid markets may have borrow fees of ten to fifty basis points annually. Hard-to-borrow securities, including many small-caps and some crypto assets, can have borrow fees of five to twenty percent annually or more. These fees are predictable for widely available securities but can spike unexpectedly for scarce ones, transforming a profitable short position into a losing one.

The interaction between leverage and costs creates a convexity effect. When a leveraged strategy performs well, the financing cost is a small percentage of profits and easily absorbed. When the same strategy underperforms, the financing cost becomes a larger percentage of diminished capital, accelerating the drawdown. This asymmetry means that leverage amplifies not just returns but also the sensitivity to cost assumptions. A strategy that works at two-to-one leverage may fail at four-to-one leverage not because the signal is weaker, but because the financing cost overwhelms the thinner margin.

• Require gross and net returns to be reported separately. Net returns should include all trading costs, financing charges, and operational overhead.
• Estimate total cost drag by multiplying turnover by estimated cost per round trip. If cost drag exceeds fifty percent of gross alpha, the strategy is marginal.
• Determine the strategy's capacity limit by modeling market impact at increasing order sizes. The capacity limit is the point where impact reduces net alpha to zero.
• Measure implementation shortfall by comparing paper portfolio returns to live portfolio returns. A shortfall above one percent annually is a red flag.
• Analyze the financing cost sensitivity by modeling strategy returns under different interest rate scenarios. High sensitivity to rates increases refinancing risk.
• Assess borrow cost stability for short-selling strategies. Hard-to-borrow securities can see borrow fees spike by an order of magnitude without warning.
• Compare the strategy's expected net alpha to its maximum drawdown. A strategy with a net Sharpe below zero-point-five and a Calmar below zero-point-three is unlikely to be deployable.
• Test execution quality on a pilot allocation before full deployment. Real execution often reveals costs that backtests underestimate.