Thursday, March 22, 2007

Marlyn's Trade

A Reader (Steven) asks
When one is feeling long or short on the market is it better to buy the QID/QLD (or the other ultra ETFs they have) or short it?

Example: I think the market is going down I can either buy the QID or short the QLD. Which one is better?

I enjoy these kinds of questions because there is no right answer - there is, of course, Marlyn's answer and that you can take or leave at your discretion.

First there is no right answer. I looked up the short interest for both of these puppy's and was astounded to find that, yes, folks are shorting QID, not much but they do have shorts against a "shorting" ETF. Then I saw that folks are also shorting QLD - again not a lot but the interesting thing is there are three large months that folks are short - Last July, November, and March 2007. I don't know about you but I seem to remember that there was a mini-boom from July to December followed by yet another mini-boom from November through the end of the year - so if we are looking for a contrary signal - have we found it? Only time will tell - but if it turns out the larger than normal short interest in QLD presages a boom - you heard it here first.

Now let's get back to the question. I don't short for the simple reason that I have a working set of tools for the long side, they serve me well and golf season is upon us again. I'd rather take dull or down days off than look for short opportunities so I can bat a little white ball around the course.

But I do like the idea of the QID/QLD as a pair and I took a look at some methods where you would buy them both simultaneously. What? Yes - just like some crafty old hedge fund manager type guy you would buy them proportionally to the direction of the market - I used simple numbers - 5 and 10 - for the demonstration but I'm sure you guys can work on it and get some optimums.

But here's what I did. If my sense was that the market was going up then I bought 10 QLD and 5 QID lots. If I sensed the market was turning I switched the order around. I did this in such a manner that you could get the sense of the market by 3 P.M. of any day and be positioned for the next day before 4. I used NewMoMo but you could use anything - including RSI(2) or ATR(10) or ATR(4) or a set-up or what ever you chose.

Here is a chart going back to 26 December through yesterday of the combined returns using the method above - There were 9 market turns in the period -

Do I recommend this - not on your life buster and if you use it and blow out your account zigging when you should be zagging you didn't hear it here. Actually that's the beauty of the approach - you won't blow out your account.

Now would you make as much as you would if you just went one way and not the other - no - but the method I've proposed is really low risk and if the market takes off in either direction you could always cancel one side of the trade or the other until it settled. Note that the chart reflects the fact that I held both sides through the entire test period.

So think about it - work on it on your own and maybe you will come up with the poor man's magic hedging system. If you do and you start selling it to unwary bystanders and you start making a small fortune off it - remember to call it by it's correct name - Marlyn's Trade.


Bullish Jim said...

Interesting idea.

As someone who can't watch the market all day long I've thought about the idea of going long QLD and QID simultaneously as a day trade. I guess it's sort of a bet on volatility as it would involve putting a tight stop in each and letting the one that heads in the right direction run. I'm thinking the trade could be managed with a series of conditional orders. John Carter talks about a similar strategy of going opposite directions in a currency pair and the idea has caught my imagination a little. The conditional order exits could be based on pivot points as a profit goal. Of course, in a choppy market I'd probably get stopped out in both directions.

If I'm thinking about it I'm assuming people are already doing it or it's a bad idea. :)

Marlyn Trades said...

It's a good idea but you need a good run in either direction to make it pay. I've done it with the Q's and QID and I know it works - Also you need to place at least half the trade the night before based on the way you think the market might go the next day. Not easy but you really need to get the first bounce to make it really pay.

IIO said...

Correct me if I'm wrong, but don't etf fees gradually erode the price of the etf? So, if you wanted to short the market, in theory, shorting QID (rather than buying QLD) would provide a return higher by the amount of the fee.

IIO said...

Oops, I had QLD and QID mixed up.

Marlyn Trades said...

Look all I'm doing is throwing out an idea here - I'm sure there are fifty or a hundred better ways to make the play but this is just one way and it happened to be profitable over an extremely randomly selected period.

Anonymous said...


Does this chart show % return?

The strategy made over 50% since Dec 26?


Marlyn Trades said...

Oh I wish - no that's just dollars and that is 10 shares vs 5 shares. I could have used 1 share vs 2 shares but 10 and 5 was what I picked. If anyone did do something like this let's hope they'd be using 10K shares vs 5K shares and increments thereof because you don't get much for 50 bucks these days. Now if it was 10K and 5K you would have made about 50 grand (less commissions) virtually risk free.

As I said - this is just a thought - it needs developing - I think there is something there but I don't have the time right now to work on it and hone it.

Steven said...

Marlyn - Thanks for using my comment. Since you do not short I will rephrase the quesiton as a long.

If QLD is the leveraged (2x) long, what is the better buy: if the long tools you have indicate that QLD is a buy, should one purchas QLD or sell short QID (which is another way of saying you are long QLD).

Since selling QID is just another means of saying you are long QLD your long tools shold work as you apply the long tools to QLD and act the inverse with QID, i.e, if the trend of QLD is up you sell QID and when the long tools indicate sell for QLD you buy back and cover QID.

Why may this be useful?

In theory for every dollar the QLD moves up the QID should move down as they are the mirror of each other.

But do they?


Another fine trade.