Kevin_in_GA
4,599 posts
msg #82224
- Ignore Kevin_in_GA
modified |
11/1/2009 1:03:41 PM
I have recently begun to look to volatility as a good measure of the market's likelihood to move in one direction or the other. On Friday the ^VIX went up 23.95%. This percentage increase is VERY uncommon. In fact, it has only happened 20 other times since 1/1/1999.
So what was the next day performance of the SPX?
In 16 out of 20 times, it was up the next day, by an average gain of +1.22%.
Two days out, the SPX was up 11 out of 20 times, with an average gain of +0.68%.
Perhaps most interestingly, three days out the SPX was also up 11 out of 20 times, with an average gain of +0.25%.
What this tells me is that Monday represents a great short term buying opportunity. In fact, I picked up shares of TNA right at the close on Friday.
By the way, of those 10 instances only two occurred on Fridays, with one having the SPX down fractionally on the following Monday, the other fractionally up. Therefore no "weekend bias" seems to be present.
Kevin
|
Kevin_in_GA
4,599 posts
msg #82232
- Ignore Kevin_in_GA |
11/1/2009 3:06:22 PM
Also on Friday the ^VIX went above its upper Bollinger Band(10,2) . This is a more common occurance than the 20+% rise, as it has occurred 104 other times since 1/1/1999.
So what was the next day performance of the SPX using this system?
In 64 out of 104 times (61.5%), it was up the next day, by an average gain of +0.34%.
Two days out, the SPX was up 63 out of 104 times (60.6%), with an average gain of +0.57%.
Three days out, the SPX was up 61 out of 104 times (58.6%), with an average gain of +0.29%.
By the way, of those 104 instances 21 occurred on Fridays (indicating no bias for Fridays). Of these, 17 were up the next day (81%) with the average gain on all occurrances being +0.42% after one day, +0.88% after two days, and +0.44% after three days.
|
BarTune1
441 posts
msg #82244
- Ignore BarTune1 |
11/1/2009 6:04:06 PM
Hey Kevin,
I'm happy to see any stats you have. Not quite sure how you are backtesting, but here is another suggestion. I suspect that if you filtered your results to occasions where the SPY is trading above its 200 dma vs. below its 200 dma, that the VIX (buy)signals on upward spikes will be slightly more accurate than when the SPY is below its 200 dma.
Have you considered testing downwards spikes also? That is, because I trade them too. I suspect that you get a more accurate sell signal also on a downward spike when the SPY is below the 200 dma.
|
Kevin_in_GA
4,599 posts
msg #82245
- Ignore Kevin_in_GA |
11/1/2009 6:15:51 PM
Have you considered testing downwards spikes also? That is, because I trade them too. I suspect that you get a more accurate sell signal also on a downward spike when the SPY is below the 200 dma.
++++++++++++++++
That's next. I wanted to look at the up spike, given Friday's activity.
I backtested by downloading the VIX and SPX data since 1/1/1999 into excel, and analyzing it by writing some quick formulae and IF statements. Easier than doing it from graphs.
Have not looked at 200 dma data.
|
Kevin_in_GA
4,599 posts
msg #82279
- Ignore Kevin_in_GA
modified |
11/2/2009 4:09:38 PM
OK, it's Monday and the markets are closed. How good was the VIX prediction?
Dow
9,788.68 +75.95 (+0.78%)
S&P 500
1,042.81 +6.62 (+0.64%)
Nasdaq
2,049.20 +4.09 (+0.20%)
Midcaps
662.09 +2.94 (+0.45%)
Smallcaps
298.12 -0.95 (-0.32%)
Russell 2K
560.56 -2.21 (-0.39%)
Nasdaq 100
1,672.91 +5.78 (+0.35%)
S&P 500 up, as the VIX would have predicted - higher than average, even. Based on the 2 day data post-VIX spike, tomorrow will be up from today's close as well.
|
tmanbone
124 posts
msg #82280
- Ignore tmanbone |
11/2/2009 4:43:59 PM
DOW, S&P & Nasdaq futures up in after hours trading; Russell should follow but not trading as of this post.
|
Kevin_in_GA
4,599 posts
msg #82289
- Ignore Kevin_in_GA |
11/2/2009 7:10:41 PM
More statistics - loads of fun!!!
When the VIX closed above the upper BB(10,2) on a Friday, I had reported that 17/21 times the following Monday was an up day. Today was no exception.
So what are the stats for the day following an up Monday?
In only 10 of those 17 times (59%) was the next day higher than the Monday close, but on average it was up +0.95% from the Friday close, and up +0.53% from Monday's close.
|
trendscanner
265 posts
msg #82291
- Ignore trendscanner |
11/2/2009 7:31:38 PM
I like this VIX strategy and think it's worth trading with a portion of my funds. Bought TNA at 35.60 during the AM pullback. Hopefully can sell later this week for at least a small profit. If this is an intermediate bottom, a 10 to 20% profit target is not unreasonable
Heavy volume on Friday and today, RSI divergence on daily and hourly charts. Could see it lower tomorrow or Wednesday but I think the odds are good for a profitable trade.
|
BarTune1
441 posts
msg #82293
- Ignore BarTune1 |
11/2/2009 9:42:38 PM
At 10:30 this morning a few of my postions had crossed the 5 dma (my normal exit) so I put in a few stop loss orders (fearing a potential 1st hour reversal) and luckily got out near the high. All profitable trades. Thought about buying back in in the afternoon but didn't because I was tied up with work.
Still long SSO, IYG & EWC. I anticipate a further short term move to the upside based on the variance. Hopefully, those trades will end in the green also.
Whether or not one trades this strategy using the SPYs or leveraged ETFs or options .... my initial observation was, at the very least, one should be aware of the VIX variance in timing equity trades in any circumstances.
Kevin, the only concern I had with your BB buy strategy posted in the other forum was that it was comprised of all long positions. I think you mentioned at one point (I may be mistaken) some concern in the event of a reversal in the market such that it went into a prolonged downturn and dragged all your positions (portfolio) down with it. Therefore, from a risk management perspective, I would strongly considering hedging your portfolio with puts or inverse ETFs on VIX downwards spikes. Something to think about. I am sure you are going to test the other side of the VIX spikes in time.
Cheers
|
tmanbone
124 posts
msg #82298
- Ignore tmanbone |
11/2/2009 11:43:31 PM
SSO hit the 5ma profit target intraday, if I'd have put a profit target stop in the trade, it would have triggered. Set profit target stop for tomorrow at the 5ma. Note to self: study all possible ways to protect profits in option trades i.e. target stops, trailing stops, stop losses % & price. Adapt, improvise & conquer.
|