Size Premium After Costs
The “size factor” refers to the tendency for smaller companies’ stocks to earn higher average returns than larger companies’ stocks, after adjusting for other exposures. In factor research, the size effect is often measured using a portfolio that is long small-cap stocks and short large-cap stocks, then tested across time periods. The headline result can look strong in backtests, yet the realized premium can shrink once you subtract trading costs, bid-ask spreads, market impact, and turnover drag.
Trading costs matter more for small caps because liquidity is typically lower, spreads are wider, and price impact rises when you trade larger dollar amounts. A strategy that rebalances monthly or quarterly can also face higher turnover than a buy-and-hold approach, which means costs compound over time. Even when the gross factor return looks attractive, net performance depends on how the strategy trades, how it defines “small,” and how it handles corporate actions.
One practical example: a small-cap factor index that reconstitutes on a schedule may require buying newly eligible names and selling those that moved into a larger bucket. That reshuffling can create short-term trading pressure and higher execution costs, especially during volatile periods. If you’ve ever watched a small-cap ETF’s daily spread widen during stress, you’ve seen the same friction show up in market microstructure.
Why Investors Overestimate It
Many investors treat the size premium as a single, stable number. In reality, the effect varies across time, market regimes, and definitions of size. “Small” can mean market capitalization at a point in time, a trailing average, or a rank-based cutoff, and those choices change which stocks enter and exit the portfolio.
Backtests often assume ideal execution, then subtract a generic cost estimate. Real trading rarely matches that assumption. Bid-ask spreads for small caps can be several times wider than for large caps, and market impact can grow nonlinearly with trade size. When a strategy uses screening and rebalancing, it also faces implementation shortfalls from timing, partial fills, and order-book depth.
Another dependency is the factor’s overlap with other exposures. Small-cap portfolios can tilt toward value, momentum, profitability, or leverage depending on how the universe is constructed. If the strategy is “small-only,” it still inherits correlations through sector composition and business-cycle sensitivity. That makes it harder to attribute net results to size alone, especially after costs.
Corporate actions add another layer of friction. Mergers, delistings, and share-class changes can force trades at unfavorable times. Factor models also need rules for how to handle missing prices, stale quotes, and extreme returns around events. If those rules differ between research and live trading, the net premium can diverge.
Finally, investors sometimes ignore survivorship and investability bias. Research datasets can exclude companies that went bankrupt or were acquired, which can inflate historical returns. Live strategies cannot trade delisted stocks, and they must manage the risk that a “small” name disappears after you buy it. That risk is not just theoretical; it shows up as realized losses and turnover.
Stress-Testing Net Size Returns
Model Costs With Real Inputs
Start with a cost model that reflects small-cap trading. Use at least three components: bid-ask spread estimates, estimated market impact, and explicit fees (fund expense ratios, brokerage, and taxes where relevant). For bid-ask spreads, you can use historical spread proxies from market data vendors or approximate using average quoted spread and volume. For impact, a simple approach is to scale impact with participation rate, then test a range of assumptions because impact is sensitive to order size and liquidity.
Turnover is the bridge between gross factor returns and net returns. If your rebalancing schedule implies 30% annual turnover, costs scale roughly with turnover, though the relationship depends on how trades are executed. A mild aside from implementation work: many backtests assume rebalancing at the close, but live trading often executes over a window, and that timing difference can change fills. If you’re using a tool like Python backtesting libraries (for example, backtrader or vectorbt—versions vary), confirm the timestamp alignment between signals and trades.
Check Liquidity and Capacity
Liquidity constraints can dominate net results. Estimate average daily dollar volume for the names you plan to hold, then compare it to your intended trade size. If your strategy is small-cap long/short, the short leg can also be hard to borrow, which adds financing costs and potential borrow-rate spikes. For a long-only small-cap strategy, capacity still matters because rebalancing requires buying across many names.
Use a “capacity stress” test: simulate what happens if you trade at 1x, 2x, and 5x your baseline order size. You’ll often see net returns degrade faster than gross returns because impact rises with trade size. This is where many investors get frustrated—gross factor charts look smooth, but the net curve bends downward once you add realistic execution.
Define Small-Cap Rules Clearly
Size definitions change both the portfolio and the turnover. Decide whether “small” is based on market cap rank, a cutoff, or a percentile bucket, and whether it uses current market cap or a lagged value. A lag reduces look-ahead bias but can also delay reclassification, affecting turnover and exposure. If you’re evaluating an ETF or index, check its methodology documentation for reconstitution frequency and any buffer rules around the cutoff.
Also check whether the strategy uses float-adjusted market cap. Float adjustment can exclude shares that are not realistically tradable, which changes the investable universe. In some index methodologies, the treatment of multiple share classes and foreign listings can affect which names qualify. Those details rarely show up in a simple “small vs large” chart, yet they can influence net returns after costs.
Measure Drawdowns and Regime Risk
Even if net returns remain positive after costs, the path matters. Small-cap factors can underperform for extended periods, especially when risk appetite shifts toward large, liquid growth stocks. Track rolling returns and drawdowns rather than relying on a single full-sample average. A practical method is to compute rolling 12-month and 36-month net returns under your cost assumptions and compare them to a large-cap benchmark.
Regime risk also interacts with financing and liquidity. In stress periods, spreads widen and trading becomes more expensive, which can worsen net performance exactly when the factor is already under pressure. If your cost model uses average spreads, rerun the analysis with wider-spread scenarios, such as using the 75th or 90th percentile spread over the last few years. That approach often reveals how fragile a premium can be when liquidity dries up.
Educational Case Examples
Scenario A: Monthly Rebalance, Moderate Turnover. A researcher tests a long-only small-cap screen with monthly rebalancing. The gross size tilt looks positive over 10 years, but the portfolio holds many names with low average volume. When the researcher subtracts a spread estimate of 0.20% per trade and assumes 25% annual turnover, the net premium shrinks to a level close to the benchmark’s tracking error. The researcher then adds a market impact term that scales with participation rate and finds net results become sensitive to the impact coefficient, which is consistent with real execution constraints.
Scenario B: Long/Short Factor, Borrow and Impact Costs. An investor evaluates a small-minus-large long/short factor. The backtest assumes shorts can be executed at the same cost as longs. In live conditions, the short leg faces borrow constraints and variable financing rates, and the strategy trades during reconstitution windows. After adding estimated borrow costs and a conservative impact model for both legs, the net factor return becomes less stable across subperiods. The investor observes that the factor’s worst months coincide with wider spreads and higher borrow costs, which complicates attribution to size alone.
Checklist For Net Premium
| Decision Point | What To Verify | Why It Changes Net Returns | What To Test |
|---|---|---|---|
| Cost Model | Spreads, impact, fees, taxes | Small caps trade with wider spreads and higher impact | Use multiple spread percentiles and impact coefficients |
| Turnover | Rebalance frequency and turnover estimate | Costs scale with trading frequency | Run sensitivity at 15%, 30%, 50% turnover |
| Size Definition | Cutoff, rank, float-adjustment, lag | Different rules change membership and turnover | Compare two plausible size methodologies |
| Liquidity & Borrow | Average volume, borrow rates (if short) | Execution frictions rise during stress | Stress with wider spreads and higher financing costs |
Step-by-step checklist for a personal evaluation: (1) pick a size definition that matches the product or strategy you’re considering, (2) estimate turnover from the methodology or from historical holdings, (3) add a conservative cost model that includes spreads and impact, (4) test rolling net returns and drawdowns, and (5) compare net results to a realistic alternative benchmark, not just a broad market index.
Common Mistakes To Avoid
One frequent mistake is subtracting a single flat cost rate from the gross factor return. Small-cap trading costs vary by liquidity, volatility, and time of day, so a flat rate can understate the cost during the exact months when the factor underperforms. Another mistake is ignoring the timing of signal formation and trade execution, which can create look-ahead bias or misalign reconstitution dates.
Investors also overfit. If a backtest uses many parameter choices—size cutoff, rebalance frequency, and filters for liquidity—then the net premium can disappear when you hold parameters fixed out of sample. A mild frustration here: it’s easy to “tune” a strategy until it looks good, then forget that live trading will face different spreads and market impact.
Another error is treating factor returns as guaranteed. Even with costs modeled, factor premiums can be negative for long stretches. If you only look at the average return, you may miss the probability of large drawdowns and the risk that you’ll abandon the strategy at the wrong time. Finally, investors sometimes ignore survivorship and investability bias when using historical constituents, which can inflate the apparent net premium.
FAQ
What does “small-cap premium” mean?
It refers to the average return difference between smaller and larger stocks after accounting for other factors in a research framework, often measured with a long-small/short-large portfolio.
Why do trading costs reduce the size effect?
Small caps usually have wider bid-ask spreads and higher market impact, so frequent rebalancing and larger order sizes can turn a positive gross premium into a smaller or negative net result.
How can I estimate turnover for a small-cap strategy?
Use the strategy’s rebalancing schedule and changes in holdings, or estimate turnover from historical holdings data if available; then test how net returns change under plausible turnover ranges.
Do small-cap factors overlap with value or momentum?
Yes. Size-based portfolios often tilt toward value and other characteristics because smaller firms’ fundamentals and trading behavior differ from large firms, so “size” can be a proxy for multiple exposures.
What net performance should I compare against?
Compare net returns after costs to a benchmark that matches your investability constraints, such as a broad small-cap index or a relevant large-cap benchmark, and evaluate rolling returns and drawdowns.
Author's Insight
Size-factor research often reports a premium in gross returns, but net outcomes depend heavily on implementation details. The most defensible approach is to treat trading costs as a range rather than a single number, because spreads and impact vary with liquidity and volatility. When evaluating a small-cap strategy, I focus on turnover, execution assumptions, and how the strategy defines “small,” since those choices drive the gap between backtests and live results. If you want a quick sanity check, compare the strategy’s implied trading intensity to the liquidity profile of its holdings, then stress the cost model during high-volatility periods.
Key Takeaways
- Small-cap premiums can shrink after trading costs because small caps tend to have wider spreads and higher market impact.
- Net results depend on turnover, execution timing, and the exact definition of “small,” not just the historical average factor return.
- Stress-test spreads, impact, and (for short strategies) borrow and financing costs, then evaluate rolling returns and drawdowns.
- A strategy that looks strong in a gross backtest can become fragile once you model realistic implementation frictions.
