This Thread has 3 Parts for a Total of 18 Minutes

Robert Paulson, CFA has just finished a trade in his portfolio for a purchse of PSSC (Paper Street Soap Co.). The trade took several days to complete and Robert wanted to calculate how much the trading process cost through the use of an implementation shortfall calculation. The original trade was put on for 1000 shares with a limit price of $30.00 right after the close on Moday when the stock closed at $30.05. No shares were purchased on Tuesday and the closing price ended at $30.25. Robert revised his limit before Wednesday’s open to $30.30 and 900 shares were purchased at that price with commissions of $25. At the end of the day the stock closes at $30.50 and Robert cancelled the remaining trade.

A. Identify & Describe the different parts of implementation shortfall (2 minutes)

B. Calculate each part of the implementation shortfall strategy and Determine what the total Implimentation shortfall for this trade is. (10 Minutes)

C. List 4 advantages and 2 disadvantages to using an implimentation shortfall strategy. (6 Minutes)


  • i. Explicit Costs: Commisions, Taxes etc.
  • II. Market Impact Cost: Change in Transaction Price due to the order
  • iii. Missed Trade Opportunity Cost: Unrealized Profit/Loss due to failure to complete the trade in time.
  • iv. Delay Cost: Cost due to inability to fulfill the order due to size / liquidity of markets.


  • i. Explicit Costs: 25/1000 = 25 bps.
  • ii. delay cost = (30.25 - 30.05)*900/1000 = 0.180 = 180 bps
  • iii. Realized Profit / Loss = (30.30 - 30.25) * 900/1000 = 0.045 = 45 bps
  • iv. Missed trade opportunity Cost = (30.50 - 30.05) * 100/1000 = 0.045 = 45 bps


  • Return on Paper Portfolio = 1000 * 30.50 - 1000 * 30.05 = 450
  • Real Portfolio = 900 * 30.50 = 27,450
  • Actual Cost of Portfolio = 30.30 * 900 = 27270
  • Commissions = 25
  • Total = 27270 + 25 = 27295

Net gain on portfolio = 27,450 - 27,295 = 155

Implementation Shortfall = 450 - 155 = 295

Add our costs: 25 + 180 + 45 + 45 = 295 bps

C. Advantages

  • Links trading to portfolio manager activity (to cost of investment ideas)
  • Recognizes tradeoff between immediacy and price
  • allows attribution of costs
  • can be built into portfolio optimizer to reduce turnover and increase realized performance
  • cannot be gamed


  • requires extensive data collection and interpretation
  • imposes an unfamiliar framework on managers.


realized: cost for executing trade at unfavorable price

delay: cost for delaying the trade which ended up at unfavorable price.

M:TOC: opportunity cost on shares that were canceled and should have earned some profit





decomposes cost into components

easily calculated

calculation procedute kills time

you can show off this method to fellow traders


not many investor know about this

if benchmark price is changed, result changes

lol nice cpk, my number is totally off. I know where i missed. worth a shot tho.

I think the denominator in the calculations is incorrect. It should be the benchmark price times the trade order size i.e. $30.05*1000 = $30,050.

(Also Otherwise for example 25/1000 would be 250 bps , not 25 bps , which is too large )

My calculation is

Commissions: 8.3 BPS

Delay costs 59.9 bps

Realized gain / loss 15 bps

Missed trades 15 bps

Total = 98.2 bps

Commissions = $25/( $30.05*1000)= 8.3bps

Delay = ($30.25-$30.05)*900/( $30.05*1000) = 59.9 bps

Realized Gain/loss = ($30.30-$30.25)*900/($30.05*1000) = 15 bps

Missed Trades costs = ($30.5-$30.05)*100/($30.05*1000) = 15 bps

Or total of 98.2 bps

I am “all in” with janakisri

accepted my sol was wrong … missed the 30.05 on the denominator for all.

how does the actual vs. paper portfolio comparison work with this price? (for the 98.2 bps?)

I saw that CPK, I think it’s because you din’t use the denominator anywhere - even your paper portfolio return. your answer was the actual $ loss, not the percent.

If you take your answer of .295 and divide it by 30.05 you get ~.0098. I think as Jana mentioned you also had some decimal places off. I bet it would count for partial credit though.

How about this :

Paper Portfolio Made (30.50-30.05)*1000 = $450

Actual made : ($30.5-$30.3) *900 Gains - $25 commissions = $155,

So Shortfall was $450-$155 = $295

in basis points = $295/($30.05*1000) = 98.2 bps

This is just a check.

Value of the paper portfolio at time 0: 30.05*1000=30,050 Value changed of the paper portfolio: (30.50-30.05)*1000=450 Value changed of the traded shares: (30.50-30.30)*900-25=155

Implementation cost = (the gain of paper portfolio - the gain of traded shares)/(papaer portfolio at time 0) = (450-155)/(30.05*1000) = 0.009817 = 98.17 bps




I almost duplicate what jana did. :slight_smile: …I’ll practice this kind of questions until it’s very hard to make mistakes and finish fast. One sanity test: % = (/)*(shares/shares)

Prophets - The first rule of this thread is not to talk about this thread - Good Catch!

I tried to graph this issue

Plan P Next close P Purchased P Close P


30.5 30.25 30.3 30.5

!- delay -------------------!------ relized lose----------- -!----unrelized gain-----------------!

!-----------------------------------missed trade------------------------------------------------------!

Is that easy to understand ?

@cpk - even though your answer was totally wrong (confusing $'s and bps, come on now son…), the labeling and organization made me think you knew exactly what you were talking about. Wonder how much handwriting, organization, and labeling affect the graders? FTW!

Also, why do they call it realized P&L - seems like just another delay cost to me…

Realized gain/loss may include the market impact… as for the delay, I agree with you. Since it doesn’t tell when the orders are executed during the day, it may use implementation shortfall strategy – purchasing early in the morning.

Realized p/l tries to measure the market impact , while delay measures the cost of inactivity or waiting for the trade to come to you . Slightly different attitudes to each

Realized P/L (or) Market Impact = Change in Transaction Price due to impact of trade.

Delay = Manager induced (maybe lethargy, trying to game the system … waiting for a buy order to come to his bid, or a sell order to come up to his ask … which would make a measure like Effective Spread very good - since it would be negative) - but would be caught by the IS Method, correctly.

The trade has a market impact, but it may be less than Realized Profit/Loss. The price change from decision price is caused by the trade and market movement.

I can’t find a method to calculate the market impact of a trade, which is also a part of the market. The market impact may also depend on the size & emergency of the trade.