Short QID or Long QLD?

A reader wrote (a long time ago) asking my opinion regarding whether it was better to short QID rather than go long QLD.

I've done some thinking about this and a couple of studies and we will get to the math in a minute. The most exciting news is that Proshares seems to have worked out their problems with these two ETF's and they have established the mirror image that they are supposed to have.


Even the most casual observer will see that these two funds were skewed early in their history but now they are achieving that most wonderful moment of all - symmetry. Of course that's because of the averaging that we go through in Marlyn's Curve. We would expect to see symmetry if the funds were supposed to be symmetrical.

But are the results of each fund symmetrical as well? Over 178 days during which there were 102 days when you could have bought QLD or sold QID the average returns were 1.01248% (QLD) vs 1.01286% (QID). Which is not much variance on a share by share basis - but over thousands of shares and hundreds of transactions it could add up to a significant amount of money.

But sometimes percentages are not all they are cracked up to be. Sometimes you need to go into the numbers to get the real story. And that is what I did - Instead of looking at raw percentages I looked at the dollar rate of return for each transaction. And then I summed that value and took an average across all of the transactions. Needless to say on that basis QLD came out the best ($1.02 to .70) but ... and there's always a "but" ... I failed to normalize the prices one to another - in other words I needed to make them equivalent money-wise. Once I did that here is what I found.

In the first half of the 178 day period the return was $1.201 per basis for QID vs.
$1.125 per basis for QLD. Over the entire period the return per basis for QID was 1.078 vs. 1.026 for QLD.

On Friday for example had you shorted 1.95 shares for every one share you could have purchased of QLD you would have made $2.38 vs $2.14 per transactional basis. But (another "but") that's an awful lot of math for a simple stock transaction so lets look at it another way.

Had you just shorted 1000 shares of QID you would have made $1220 and had you bought a 1000 shares of QLD you would have made $2140. But you would have had to have put up over 90 thousand dollars in the QLD transaction and a minuscule percentage of that in the QID transaction.

So my thinking is this. On days when you can be pretty sure that the market is going up (and Friday would have been a good one) shorting the QID model seems to be the better idea if you can get the shares - and that might be the only difficulty with this whole idea.

Of course if you want a technical reason to buy or short either of the products you would put up a four-minute chart of either or both of them with an EMA 90 and wait for the cross.

Here is QID on Friday -



And here is QLD -


You can see how once the stocks crossed the EMA 90 the die was cast and it was an easy trade to make either way.

Long QLD or Short QID

This remains the question - Steven asked again and I will do my best to answer -

if in a long situation which would be better - short the QID or go long the QLD?

They are supposed to be mirror images but they are not -


They are close but the problem right now is that there just isn't enough data or experience with these two funds but right now the fact is the QLD has "longer legs" than the QID.

That would suggest that anytime you have an opportunity to buy QLD vs short QID you probably should and vice versa - I.E. if the occasion suggests being long QID you probably should short QLD. As an interesting aside - the actual short interest on QID is nearly 3 times that of QLD - so we need to go to the raw figures to work it out.

Putting aside expenses and fees associated with ETFs and shorting which will change the results a bit this is what I found using 5 different "long market" periods in the past 150 days -

QLD per share total returned 13.50 and short QID per share total returned 10.03.

This bears out the visual above (QLD having "longer legs") and suggests that people who short these two stocks really don't know what they are doing because they are shorting the wrong one.

Again - the sample size is too small to say with any conviction do this or do that. All I can say is that up till now it appears a long purchase of QLD when the market goes up and a short sale of QLD when the market goes down might be the appropriate way to go. A year from now the story may change significantly.

I hope that answers the question and I'm always glad to help out.

QQQQ and QID

In our never ending search for an automated let me sit back and watch and make money method, Dogwood and I are trying a number of approaches to using the Proshares ultra-short funds along with the Q's to see if we can find a market timing system that makes sense. Dogwood is right now studying the effect of TICK on the process and I'm looking at several other methods.

Including buy both and let one fail and one succeed. The best way to do that is to buy one or the other the night before depending on your feel for the market on the next day. If you believe the market is going up the next morning buy the Q's and set a stop close below them. If you believe the market is going down buy QID and set a stop below it. Usually at 3:45 EST you shouldn't have a lot of trouble getting filled because the day traders have all gone home for the day. (Except for a few hard core who play after hours). Then the next morning at the open you buy the other automatically also with a close stop and hope that the whipsaw of the market place doesn't take them both out. It happens. I've tried it a couple of times with success each time but I don't like the suspense.

Yesterday morning provided a great opportunity to try out a pivot point based approach because the Q's opened up and QID opened down. If you want to be prepared for the next day you have to generate the pivot points before the market opens - we've discussed how pivot points are generated so I won't cover that again.

Once you have generated the pivot points for the two stocks as soon as the market opens and you can get a one day chart you should bring it up and annotate it with the pivot point, MPH, MPL, R1, S1 or as much as will fit. Some systems provide this capability built in - mine doesn't but that's OK - I would continue to annotate them by hand as I have been doing all along. I enjoy doing it and it helps me concentrate on the task at hand.

Here is what the two stocks looked like at the end of the day.





You can see how they are an exact mirror image of one another and that they hit the pivot points at the same times. Note how you could have used the pivot point (1) of both stocks to inform your purchase. Then when the QID pulled back away from R3 and the Q's pulled away from S3 you sell QID and buy the Q's.

Thus in one day you make a buck and a half on QID and another 30 cents on the Q's. Dogwood suggests that we use QLD but I don't think it has enough volume yet to make it reasonable. But if you could have got a fill in QLD around 12:15 or so you would have made about a buck on the transaction.

I'll stick with the Q's for the counter trade in this instance until QLD becomes more popular.

A little memory trick so that you won't be buying the wrong ETF - QLD - "Long Double Q's", QID - "Inverse Double Q's".