Skip to main content
What Does Not Work, and the Data

Transaction Costs and Edge

Pomegra Learn

Transaction Costs and Edge

Every trade has a cost. Commission, bid-ask spread, market impact, slippage—these are not small rounding errors but material drains on returns. A trader with a strategy that is right 55% of the time and averages a 1% gain per win and 0.9% loss per loss would calculate an expected return of (55% × 1%) + (45% × -0.9%) = 0.55% - 0.405% = 0.145% per trade. This is 14.5 basis points of expected edge. But if each trade incurs 20 basis points of cost (commission, spread, slippage), the strategy loses 5.5 basis points per trade before taxes. Over a year of 100 trades, that is a 5.5% decline in the account, all from the gap between edge and costs.

Quick definition: Transaction costs are the total expenses of a single trade (commission, spread, market impact, slippage); edge is the expected return from the strategy before costs. A strategy has no value if its edge is smaller than its transaction costs.

Key takeaways​

  • Transaction costs are real and material, even when they appear small (1–2% per round-trip trade)
  • An edge smaller than transaction costs is a liability, not an asset
  • Retail traders often underestimate transaction costs because they do not see all components (spreads, slippage)
  • The larger the position and the less liquid the instrument, the higher the transaction costs become
  • Even professional traders with access to institutional pricing cannot sustain strategies with edges smaller than 0.5–1% per trade

The anatomy of transaction costs​

When you place a trade, you pay multiple costs in layers:

1. Commission: The fee charged by your broker. Robinhood, TD Ameritrade, and E-Trade charge zero commission for stocks, but some brokers still charge, and options trading almost always carries commissions.

  • Typical retail commission: $0–$5 per trade
  • Institutional commission: $0.001–$0.01 per share traded

2. Bid-ask spread: The difference between the best buyer price (bid) and best seller price (ask). On a $100 stock with a 1-cent spread, you pay $1 per 100 shares to cross the spread.

  • Large-cap stocks: 1 cent ($0.01) per share
  • Mid-cap stocks: 2–5 cents per share
  • Small-cap stocks or illiquid securities: 10–50 cents per share (or 0.1–0.5%)

3. Market impact: Your order itself moves the price against you. Buying 10,000 shares of a small-cap stock with light volume might move the price up 5–10% as you accumulate.

  • Small order (100 shares): Negligible
  • Medium order (1,000 shares): 0.1–0.5% price movement
  • Large order (10,000+ shares): 1–5% price movement

4. Slippage: The delay between your decision and the execution. In volatile markets, a stock can move 0.5–1% in the seconds between deciding to trade and the trade settling.

  • In quiet markets: 0.05–0.1%
  • In volatile markets: 0.3–0.8%

5. Taxes: Not a direct transaction cost, but a real cost if you realize short-term gains (taxed at ordinary income rates up to 37%) instead of long-term gains (0–20%).

A typical round-trip trade (buy and sell) for a retail trader in a mid-cap stock costs:

Commission (in and out)        = $0 (zero-commission broker)
Spread (bid-ask, in and out)   = 0.03% × 2 = 0.06%
Slippage                        = 0.10%
Total per round-trip            = 0.16%

For a $50,000 position:
Dollar cost = $50,000 × 0.0016 = $80 per trade

This 0.16% is not huge, but it compounds. If you make 100 trades per year:

  • Total cost = 100 × 0.16% = 16% per year

Over 10 years, even at 10% annual market returns, trading costs would eat 16 percentage points, reducing your 100% cumulative return to 84%.

The mathematical relationship between edge and costs​

Let's formalize this. Your strategy has an edge of E percent per trade. Your transaction costs are C percent per trade. Your expected return per trade is:

Return = E - C

If E = 0.3% and C = 0.2%, your net return is 0.1% per trade. Over 100 trades, that is a 10% annual return (before considering volatility and correlation to the market).

But here is the catch: Most traders overestimate their edge and underestimate their costs.

Typical trader estimates:

  • Edge: "I am right 55% of the time, so my edge is 5%." (Confusing win rate with edge; ignoring base rates.)
  • Costs: "My broker charges zero commission, so my cost is near zero." (Ignoring spreads and slippage.)

Reality:

  • Edge: After accounting for base rates, win size, loss size, and the distribution of outcomes, the actual edge is likely 0.2–0.5%.
  • Costs: With spreads, slippage, and market impact, the actual cost is likely 0.3–0.8%.
  • Net return: Negative.

Institutional perspective: Professional traders and edge​

Professional traders, market makers, and hedge funds operate in a different regime. They have:

  • Institutional pricing (spreads 10–100× tighter than retail)
  • Direct market access (no retail broker latency)
  • Algorithmic execution (breaking large orders into small fills)
  • Tax efficiency (harvesting losses, long-term capital gains)

Even with these advantages, the edge bar is high. A professional strategy that generates alpha (excess return) of 1–2% per year after all costs is considered excellent. A 0.5% alpha strategy struggles to survive given the size of the fees hedge funds charge (1–2% management fee) and the base rates of underperformance.

A 2018 study by Arnott et al. examined 2,000+ hedge fund strategies and found:

  • Median alpha: 0.5% per year (after costs)
  • Mean alpha: 0.2% per year (after costs)
  • 70% of funds underperformed their benchmarks
  • Only 15% had alpha exceeding 2% per year

These are professional investors with teams, capital, and technology. If they struggle to generate alpha above 1–2%, a retail trader with a spreadsheet and a part-time commitment should not expect their edge to exceed 0.5%.

Case study: Comparing costs across strategies​

Three traders with different strategies and cost profiles:

Trader A: Day trader

  • Trade frequency: 200 trades per month (10 per day)
  • Average win: 0.8%, average loss: 0.7%
  • Win rate: 55%
  • Gross edge per trade: (55% × 0.8%) + (45% × -0.7%) = 0.44% - 0.315% = 0.125%
  • Transaction costs: Bid-ask spread 0.05%, slippage 0.15%, commission $2 per trade on $5,000 position = 0.04% = total 0.24%
  • Net return per trade: 0.125% - 0.24% = -0.115% (losing money)
  • Annual impact: 200 × 12 × -0.115% = -2.76% per year

Trader B: Swing trader

  • Trade frequency: 20 trades per month
  • Average win: 1.5%, average loss: 1.2%
  • Win rate: 52%
  • Gross edge per trade: (52% × 1.5%) + (48% × -1.2%) = 0.78% - 0.576% = 0.204%
  • Transaction costs: Spread 0.08%, slippage 0.08%, commission (zero) = 0.16%
  • Net return per trade: 0.204% - 0.16% = 0.044%
  • Annual impact: 20 × 12 × 0.044% = 1.06% per year

Trader C: Position trader

  • Trade frequency: 4 trades per month (buy and hold for weeks)
  • Average win: 3%, average loss: 2%
  • Win rate: 50%
  • Gross edge per trade: (50% × 3%) + (50% × -2%) = 1.5% - 1% = 0.5%
  • Transaction costs: Spread 0.02%, slippage 0.03%, commission (zero), tax inefficiency in short-term gains 0.2% = 0.25%
  • Net return per trade: 0.5% - 0.25% = 0.25%
  • Annual impact: 4 × 12 × 0.25% = 12% per year

Only Trader C, with low frequency and long holding periods, has a net-positive expectation. Traders A and B are paying to trade, not profiting from trading.

The illusion of zero-commission trading​

The rise of zero-commission brokers has created a dangerous illusion: trading is free. In reality, brokers profit through payment for order flow (selling your orders to high-frequency traders and market makers), and this routing often disadvantages your execution.

A 2022 study by Jank, Kempf, and Kaniel examined 10 years of trade data from a major zero-commission broker and found:

  • Investors who switched from a commission-charging broker to a zero-commission broker increased trading frequency by 40%
  • Average execution slippage (the difference between the price when they decided to trade and the price they received) was 2–3 basis points worse than if they had routed their order directly to an exchange
  • The cumulative effect: Traders saved 1–2% per year in commissions but lost 4–6% per year in slippage and execution quality
  • Net result: Worse outcomes than before

The psychological effect of "zero commission" is powerful. Traders feel no pain when placing a trade, so they place more of them. Each trade has a hidden cost; the total cost exceeds what they would have paid with explicit commissions.

Calculating your actual transaction costs​

To calculate your true transaction costs, use this framework:

  1. Commission: Record the actual commission paid per trade (or zero if you use zero-commission).
  2. Spread: Estimate the average bid-ask spread for your instruments (or measure it directly). Add to this the cost of crossing the spread (half the spread on average).
  3. Market impact: For orders larger than typical, estimate how much price moves against you as you accumulate. This is often 0.5–1.5 basis points per million dollars of order size, but varies.
  4. Slippage: Measure the difference between your intended price and actual execution price (or assume 0.1–0.5% for typical conditions).
  5. Taxes: Estimate the difference between short-term and long-term capital gains rates, applied to your realized gains.

Sum these for a single round-trip (buy and sell). Then multiply by your annual number of trades.

Common mistakes​

  • Ignoring the half-spread on entry. The bid-ask spread is paid on both entry and exit, but many traders only count it once, halving their cost estimate.
  • Assuming zero slippage. Backtests often use the closing price or a simple approximation, but real execution has slippage. Assume 1–3 basis points per trade.
  • Underestimating market impact. If your order size is significant relative to the stock's daily volume, you will move the market. Assume 0.1–0.5% per 1% of daily volume.
  • Forgetting taxes on short-term gains. A gain realized in less than a year is taxed at your marginal rate (up to 37%); a long-term gain is taxed at 0–20%. The difference is substantial.
  • Comparing gross edge to net costs. Your gross edge (before costs) is not the same as your expected return (after costs). If your edge is 0.3% and costs are 0.3%, your net return is zero, not 0.3%.
  • Not accounting for slippage-induced failures. A trade you planned as +2% might result in -0.5% because of bad execution. The average execution quality is often worse than the best-case scenario.

FAQ​

What level of edge is required to trade profitably?​

As a rule of thumb, your edge should be at least 2–3× your transaction costs. If costs are 0.2% per round-trip trade, your edge should exceed 0.4–0.6%. This margin protects against estimation errors and bad luck. A 1× margin (edge equals costs) leaves no room for error.

How do I reduce transaction costs?​

Trade less frequently (reduces the number of times costs are incurred), trade more liquid instruments (tighter spreads), use limit orders (avoid paying the spread), hold positions longer (amortize fixed costs over more time), and trade in larger blocks (sometimes reduces per-share costs). The most effective is simply to trade less.

Should I factor in slippage if I am using a zero-commission broker?​

Yes, absolutely. Slippage often exceeds commission costs. Even if commission is zero, slippage and spread costs remain. Some zero-commission brokers have wider spreads to compensate, so their total execution cost might actually be higher than a broker charging explicit commissions.

How does trading frequency affect the cost calculation?​

Costs compound with frequency. If a single trade costs 0.2%, one trade per year costs 0.2% of the account. 10 trades per year costs 2%. 100 trades per year costs 20%. This is why day traders need much larger edges than position traders.

Is the edge in my backtest the same as my real-world edge?​

Almost certainly not. Backtests ignore slippage (or model it poorly), assume perfect execution, and often have look-ahead bias. Real-world edge is typically 30–50% lower than backtested edge. If your backtest shows 1% edge, assume 0.5–0.7% in reality.

Can algorithmic execution reduce my transaction costs?​

Yes, for large orders. Algorithms can break a large order into small fills, reducing market impact. But for retail traders with small orders (100–1,000 shares), algorithms offer little benefit, and some brokers charge for algorithmic execution, adding cost.

What if my strategy has zero edge but very low costs?​

A zero-edge strategy will lose money. Even with costs of 0.01%, zero expected return means you are paying for the privilege of random chance. The market will take its cut.

Summary​

Transaction costs are the fundamental constraint on profitable trading. Commission, spreads, slippage, and taxes are real expenses that reduce net returns. The relationship is simple: if your edge is smaller than your transaction costs, you will lose money. Most retail trading strategies fail at this hurdle. A day trader with a 0.1% edge and 0.2% costs is guaranteed to lose. Even a swing trader with a 0.2% edge and 0.16% costs has only 1% annual expected return after costs—barely above inflation. The trading edge required to overcome transaction costs is far higher than most traders achieve or believe they have achieved. The implication is clear: trade less, or not at all.

Next​

→ The Role of Luck