I’ve been reading a lot over the past week of people talking about the overbought nature of the market. Now, they may be right or wrong in this particular case, but I was curious to see how the % of stocks over a 50 day moving average worked in terms of registering “overboughtness”. So, I set up a simple test. I first created my own % of stocks above a 50-day moving average using the SPX. Next, I setup a test where I shorted the SPY when the % 50-day went from above 80 to below 80 over two days – essentially a cross overover. I shorted the SPY on the open of the day after the event, and then I covered it 5 days later on the close. I used data on the SPY from 1989 on.
My conclusion from these statistics is that using the % over a 50 day moving average is not very good at indicating a top.
Now, there are some natural issues with this approach – the chief one being survivorship bias in the index – I can’t currently calculate the index including stocks that may have been delisted – but the initial study does not look promising.
Over the next few weeks we’ll explore this idea a bit more and see if there are any workable systems based off of this basic idea.
Statistics:
| Net Profit % | -24.88% |
| Exposure % | 8.14% |
| Net Risk Adjusted Return % | -305.62% |
| Annual Return % | -1.41% |
| Risk Adjusted Return % | -17.28% |
| All trades | 88 |
| Avg. Profit/Loss | -28.28 |
| Avg. Profit/Loss % | -0.30% |
| Avg. Bars Held | 6 |
| Winners | 37 (42.05 %) |
| Total Profit | 3886.3 |
| Avg. Profit | 105.04 |
| Avg. Profit % | 1.50% |
| Avg. Bars Held | 6 |
| Max. Consecutive | 4 |
| Largest win | 287.1 |
| # bars in largest win | 6 |
| Losers | 51 (57.95 %) |
| Total Loss | -6374.8 |
| Avg. Loss | -125 |
| Avg. Loss % | -1.60% |
| Avg. Bars Held | 6 |
| Max. Consecutive | 12 |
| Largest loss | -469.7 |
| # bars in largest loss | 6 |
There’s a big difference – HUGE – between a good short and a poor long.
Try looking at relative performance of being long during different levels of “overboughtness” from a 50 dma breadth perspective. It may be useful as a cue to lighten up winners as opposed to a cue to go short …
Agreed – I think that’s a great area for further exploration – such as dropping equity size in, say, half, when the 50 dma goes above 70. I’ll try that out this weekend. Thanks for the idea.