TO BUY OR NOT TO BUY?

I have been studying this idea now for a week or so, buying into severely oversold conditions when still in an uptrend. If you test out buying stocks at over valued levels when they are above the 200 day moving average and then exiting on some type of strength, this does test out very favorably. The problem of course is that oversold can get more oversold etc.. As you can see on the above chart, we were oversold at the first, second, third, and fourth arrows. How can we ever know when to go after a trade like this?

The answer that is put forth by Larry Connors in his book, who is a great trader, is that you leg into these types of trades. I have been doing one with the Naz that I have been showing in here, the QQQ, just to do one live for all to watch. So far there are 2 units on the trade. If today were to close lower, a third would be added.

Here is the biggest problem with this approach, and it is a BIG ONE. How do you determine position size when trading like this? I have tested out all kinds of different entry techniques in these situations, buying above the high, buying on a limit below, as well as many other ideas. None of them test out better than simply buying on weakness and continuing to do so as the decline continues. On top of that using stops also negatively impacts the results dramatically. The question then remains, if we buy into a sharp decline like this relying on the overall trend to turn it around for us, how do we determine position size? The biggest problem with this is that when there is a trend change, you get clobbered on this approach not using a stop of any kind. It pays off because the winning percentage is very high, so you make constant small amounts, occasionally a big one, then rarely, but they do happen, a big loss.

This is an ongoing project for me, so I do not have the exact answer yet, but my current thinking is along two lines. First, don't take these buys when there is an overall downtrend in the Bernanke's or there appears to be a sell setting up with them. Second, eyeball what you think the max loss per share might be, then back your way into your number of shares that way.

The first is always tricky, because we have had an absolute ton of false sell signals in the indexes. By sell signal here I mean a definitive breakdown type of situation, not a wiggle. I am not saying here that if you see a head and shoulders on a 15 minute ES chart that is a sell, to not take buys in stocks, that would be ludicrous. The second part is a judgement call, as to where you think it could fall to before you know the trend is changed. This is a work in progress, if I come up with anything I think resolves all of this I will let you know.

In the mean time, the QQQ trade I made will be exited on the close today if it stays where it is now for a profit.

On to other business, I have orders in again that are highly unlikely to be filled to short Bonds today below yesterdays low. This is a nice trap pattern if it were to reverse today.

Once again we have potentially if this were to take out yesterday's low, a confirmed trap pattern. It does not appear too likely this will happen but who knows? There was a comment made about Bonds and Stocks both going up and whether or not one of those is wrong. In the old days, buying stocks when Bonds rose was a great thing to do. As much as I do also watch these inter market correlations, it is best to ignore them when determining direction. I think they do need to be considered in position sizing, but often you miss a good move because the correlation slips in and out of it's place while the markets move along.

The most important correlation that is in place now is that of the Es and the DX, the rest I don't think matter too much.

Have a great holiday