The Opportunity Cost Of Unfilled Orders
The cost that only shows up when a trade never happens at all — a limit order that never fills, or an order canceled before completion, leaving a decision that was never actually implemented.
Prerequisites: Delay Cost And The Decision Price
A manager decides to buy 100,000 shares at $40.00, places a passive limit order at $40.00 hoping to avoid crossing the spread, and the stock simply runs away — by end of day it's at $41.50 and only 30,000 shares have filled. The other 70,000 shares were never bought at all. Measuring cost only on the shares that actually traded misses the largest problem entirely: the 70,000 shares that should have been bought, per the original decision, and now would cost $1.50 more each if bought at all. This is opportunity cost in the implementation-shortfall sense — the cost of a decision that was never fully implemented, not the cost of a bad execution on the part that was.
Why unfilled size is a cost, not a non-event
It's tempting to treat an unfilled order as simply "nothing happened" — no trade, no cost. But the original investment decision assumed the position would exist; if the strategy behind the decision was sound, failing to establish 70% of the intended position means missing 70% of whatever the strategy was supposed to capture. Perold's implementation shortfall framework treats this explicitly: for any portion of an order left unfilled, the cost is computed as if that portion had to be closed out at the prevailing price at the end of the measurement period, versus the original decision price — a paper cost, but one that reflects the real economic consequence of the decision not being fully carried out.
This cost cuts both ways depending on direction: an unfilled buy order in a rising stock is expensive because the position you don't have is now more valuable than it was; an unfilled sell order in a falling stock is expensive because the position you're still holding has lost value you were trying to avoid. Either way, being too passive to protect against small amounts of market impact can produce a much larger cost if the price simply moves against the un-executed intention.
Worked example: quantifying the unfilled portion
Continuing the example above: decision price $40.00, order for 100,000 shares, only 30,000 fill (at an average price of $40.05, say, reflecting modest impact on the filled portion), and the stock closes the measurement period at $41.50. Cost on the filled portion:
i.e. $1,500. Opportunity cost on the unfilled 70,000 shares, valued at the $41.50 closing price versus the $40.00 decision price:
i.e. $105,000.
The unfilled portion's opportunity cost — $105,000 — dwarfs the $1,500 cost on the shares that actually traded, by a factor of 70. A transaction cost report that only looked at the 30,000 executed shares would show this as a cheap, well-executed trade, when in fact the passive strategy of resting at the limit price cost the fund the overwhelming majority of the value the original decision was trying to capture.
What this means in practice
Opportunity cost is the reason execution algorithms don't universally default to maximally passive strategies even though passive orders minimize market impact — being too cautious about impact risks a much larger cost if the market simply moves on without you. Any complete transaction cost analysis has to account for unfilled size explicitly, valuing it against the decision price and a later reference price, or it will systematically understate the true cost of overly passive trading.
A common mistake in TCA is reporting execution quality only on the shares that actually traded, silently excluding the unfilled portion of partially filled orders. This flatters passive strategies that fail to complete and can make a genuinely costly decision look cheap.
Further reading
- Perold, 'The Implementation Shortfall', Journal of Portfolio Management, 1988