Implementation Shortfall - Delay Costs

[mildeng]

Hello,
Does anyone have an intuitive way of calculating delay (slippage) costs when breaking down the implementation shortfall method? I just cannot seem to get this to stick in a logical way… actually have found it easier calculating the entire shortfall overall and then subtracting each other component to arrive at this :/
As an example, I’m looking at Reading 39 p. 52 Q11(B) - method they use to calculate it (particularly a zero figure for the ‘Monday’ period I’m finding confusing)
Thanks

 

[janakisri]

I mentally think of a moving market price to measure delay cost. The delay cost is always equal to :
delay cost=( moving market price - benchmark price) *shares traded/( total shares * benchmark price )
Initially the moving price is the benchmark price . Each time there is a closing price ( i.e. a day completed ) there is a new market price to be used ( my mental moving market price)
If some trades are completed within the first day the delay cost is zero for those trades , because you did not encounter a exchange closing price. After the first day is done , the new moving market price becomes the close of the first day . If trades complete the third day you use the second day’s close as the moving market price.
Of course goes without saying that the market impact cost is the other leg . i.e. the trade price - the moving market price.
market impact cost=( trade price - moving market price ) *shares traded/ ( total shares * benchmark price )
If you add up these two, you get the value :
delay+impact=( trade price - benchmark price ) *shares traded /( total shares * benchmark price )
add commissions and missed trade opportunity costs and it should equal benchmark gain/loss - portfolio gain/loss