Timing model review

Dedicated readers will see a section in my Total market view posts, usually near the end, that I don't say much about called Timing. You can also search keywords for "Timing model" and see a few dedicated posts before I shifted to a more condensed format. 

This post is a review of the timing dates - all of which were published well in advance. Maybe members of the math, statistics and back-testing crowd who happen to be reading this blog can give me some official numbers on whether the results are random, or much too unusual to be pure chance.

Here is an SPY chart with all timing dates marked. Most of the time I did not specify bias of high or low, and simply viewed them as turn possibilities. Sometimes were was a window of a few dates clustered. So, single green lines mean 1 day; a green followed by a red means green was the start of the period and red was the end. Most of the windows were 2-3 trading days, with just a couple going beyond that.

From 2/11 when I started this project through end of October, I have listed 52 trading days (if I counted correctly) as higher chance of turn areas. Over the same period there have been 184 trading days. This means I have selected 28% of all trading days as higher than likely turns. But if we extend -1 & +1 from each day, then obviously it extends quite a bit, and maybe it is just chance.

From other view, though, I have selected 23 1-3+ day periods where we are more likely to see a turn. 

In these 52 days & 26 periods we have:

2/11-15 key stock index low of year, and oil low of year both 2/11
4/22 stock trading high -2 days arrived on 4/20
6/7-9 major stock high in middle of window 6/8
6/28 major stock low on 6/27 -1 day
8/19-22 major stock high (SPX high of year) 8/23 +1 day
9/12 medium stock index low of September exact
9/23 key stock index lower high 9/22 -1 trading day
10/7-10 key stock index high, and NDX high of year thus far, on 10/10

Then several days turned out to be minor highs or lows. A few were pure misses. 

I did not get the 5/19 stock index pullback low, 7/6-8 GLD & TLT high, or the 9/7 lower high in stocks. But I'd say all the other important turns on SPY have happened on or near timing dates.

This method cannot really be back-tested. It is nothing to do with what most of you are thinking. Not trading day or calendar day counts. Nothing Fibonacci. And remember, aside from starting on 2/11 I have usually published these days near the end of the month for the following month. That means all advance. Yup, I used to do this for a hedge fund too :) If I had more time, then it is possible to get direction (ie high or low) in addition to timing the turns. 

So if you are paying more attention now, the next date for October is 10/19. 

Maybe now some of you are starting to get it: pivots + other technicals + valuation & fundamental basics + sentiment + timing = very high chance of being on right side of the market; having exposure to leaders; avoiding, hedging or shorting what is going down; and having high reward & low risk opportunities for leveraged trading. Any funds out there seeking better performance to retain their AUM?