Is it possible that implementation shortfall is negative?
No…
Yes neg IS is good!
All implicit IS components can be positive or negative costs while explicit IS is always positive costs.
Bruh… the BEST case scenario (in lala-land) is that the very MINUTE you decide you want to buy 200,000 shares of UVXY @ $35.00 is your order INSTANTLY being filled at EXACTLY $35.00
There’s no such scenario where your ACTUAL portfolio actually beats the HYPOTHETICAL portfolio - because the hypothethical portfolio represents the absolute best case scenario, it can’t be improved on.
No… the best case scenario is that when you buy 200000 shares at $35.00, there is negative news that takes the shares to $30. Your execution price is much better than your DP/BP! That may be an extreme example, but I’ve executed market orders and gotten favorable pricing many times 
It’s also very conceivable your realized IS component is a gain rather than a loss. How does this happen? Let’s say your DP is $35 and only 10000 share executed that day. The new BP is the close of the day which is $36. But by the open the next day, negative news set and your executed price is $30! You now have a delay cost of $1 but a realized cost of -6!
. . . Damn. I know nothing anymore.
I’ve run into this problem on a mock exam… Trick one, hope its on the real thing!
Answer: Cancellation of a buy order prior to a fall in price, or cancellation of a sell order prior to a rise in price will give rise to a negative implementation shortfall.
Why???
Because price impact went in your favor.
Just the term ‘negative cost’ throws me for a loop every time. Implementation Shortfall is MUST know.
-
Paper Portfolio Gain
-
Real Portfolio Gain - Commissions
-
Step 1 - Step 2 = IS
I’m a fan of mnemonics. I think of R.E.D.O. for the breakdown.
R ealized, E xplicit, D elay, O pportunity Cost
Also, am I correct to think linearly/chronologically in time when determining what’s the DP, BP*, EP, CP? Always seems to be the case…
Yes, I do chronological time map. DP, BP*, EP, [BP*, EP*], CP.