I’ve been going back and forth on my own accounting for this situation…
It just depends what meaning you want from your calculation.
Different purposes suit different calculations.
Here’s one question that is of interest to me, as an investor:
Imagine somebody has a position in some stock, with some history of purchases and partial sales.
If they finally close that position today for proceeds of $10000, what is the final end-to-end profit on the stake since the very beginning?
After all the to and fro, what’s the bottom line at the end? Certainly a meaningful question.
That will be $10k minus whatever they have put into the the position, net, to date.
i.e., ($10k) - (the cost of all purchase to date) + (the sum of all sales proceeds to date).
The last two, the “net money in the position today”, can definitely be negative.
It’s common (and desirable!) for realized profits to date to exceed current market value.
That’s the situation I mentioned in the OP.
Anyone having reached that situation can’t have an end-to-end loss when the position is closed.
(unless perhaps they subsequently buy lots more at a high price and sell everything at a very low price)
Another commonly calculated number is the weighted average cost per share of all purchases since the share count was zero, ignoring any sales.
(That’s the cost basis for tax purposes in many countries, for example–the US seems to be an outlier).
I don’t find this number as interesting, as a position might have become big and small many times,
at a wide variety of price points through history.
Imagine I buy 100 shares at $20, sell half at $50, buy them back at $35, sell half at $65, buy them
back at $50, sell half at $80, buy them back at at $65, sell half at $95, and buy them back at $80.
I have the same 100 share position I started with at $20, and I have been doing very well overall,
but does the weighted average buy price of $45 have much meaning to me as an investor? It’s not obvious.
If the stock went to zero I’d still have an overall profit, not a loss of $45 per share, nor even $20 per share.
In this particular instance, it’s probably best thought of as two separate investments:
A long term block of 100 shares purchased at $20, and a separate bunch of short term trades showing an overall net profit of $3000.
The companion number is the weighted average sale price to date. ($72.50 in the example above).
You definitely want that to be higher than the first one.
The difference is the profit per share realized to date on whatever number of shares have been round tripped.
But other than that insight I don’t see a whole lot of use for it.
By far the most important number for me in evaluating the success of a stake is tracking the rate of return as a function of the capital at risk.
Profit is what you want, and putting capital at risk for a finite time stretch is the thing you are bringing to the party to accomplish that.
For each day, what would I lose if the price went permanently to zero? Sum all those daily figures to get a total number of dollar-days at risk.
I calculate my total profit to date, and divide that by the dollar-days at risk to get a rate of return on capital.
If that number is good, you are a good investor.
This is a good metric to see how well you are doing at dynamic position sizing: you want a lot of
capital in a position in the (probably) good stretches and little during the (Probably) bad stretches.
Digression:
Personally I find it most meaningful to calculate the “what would I lose” as “how much worse off
would I be compared to the situation the last time I resized or considered resizing the position”.
Imagine I buy a block of stock at $100.
I go on holiday for two weeks, and unknown to me the price hits $200 for most of that time.
I come back, and it’s (say) $121.
For this entire stretch I deem my capital at risk to have been $100 on each day: that’s what I brought to the party and put at risk in the position.
Viewing this interval as a standalone investment decision, $100 was my cost basis, what I decided to risk.
To me, it doesn’t make sense to view it has having had $200 at risk one week into my vacation.
I never had that $200 in hand, and I never decided to allocate it to anything, so it offers no insight into how well I have been allocating what I have in hand.
Imagine further that I sell at $121 on my return.
I’m not fond of the traditional internal rate of return calculation method, which (if you think about it)
implicitly assumes that the capital at risk was $110 half way through my vacation.
That is neither the capital I put at risk into the position for that interval, nor the market price at that intermediate moment.
Jim