What the VIX says about the next big move
The market sold off
fairly hard today as the bounce completely lost steam and rolled over.
This should come as no surprise to anyone who has read my last couple of
columns (click
here). What’s worse is that VIX levels barely budged — apparently today’s selloff didn’t worry anyone. That’s not good.
I received a large number of emails in response
to my last two columns, both of which discussed the significance of recent VIX
levels. Below is a sample of one that disagreed with my assertion that last
week’s VIX readings were signaling fear in the market.
“Rob – at these levels….the VIX below 20….you
could never assert there is any fear in this market…â€
This seems to be a common belief. Many people
think a VIX below 20 means complacency, and a VIX above 30 or 40 means fear. I
contend that absolute levels like these mean very little. What is much more
important are relative VIX readings. This is why I compared the VIX to its
10-day moving average in my studies rather than looking at any relative levels.
Why should you believe me that relative readings
are significant, while absolute readings mean little? I did two more tests so
that you may judge for yourself.
First, let’s go back and look at one of the tests
I ran in last Wednesday’s column. This is what I wrote:
Test 1
Buy the S&P 500 each time the VIX closes 12%
above its 10 day moving average.
(Do not make additional buys if you are already in a trade).
Sell 5 days later.
This simple strategy would have been successful
about 60% of the time over the test period (1991-present). The net gains over
the period would have been more than double the net losses….
Now, to test my readers’ thesis that a VIX below
20 could not signal fear, I added another condition to the above test.
Test 2
Buy the S&P 500 under the following conditions:
1) The VIX closes 12% above its 10-day
moving average.
2) The VIX closes below 20.
(Do not make additional purchases if you are already in the trade.)
Sell 5 days later.
I also looked at what happened with a VIX above
20:
Test 3
Buy the S&P 500 under the following conditions:
3) The VIX closes 12% above its 10-day
moving average.
4) The VIX closes above 20.
(Do not make additional purchases if you are already in the trade).
Sell 5 days later.
I ran each test over the last 15 years since that
mirrored the test I published last week. Here’s a comparison.
|
|
VIX < 20 |
VIX > 20 |
|
Total # of Trades |
46 |
107 |
|
Percent Profitable |
59.1% |
61.7% |
|
Profit Factor |
2.53 |
2.19 |
Winning trade percentage and profits were
basically the same whether the VIX was trading above or below 20. This seems to
support my contention that absolute values mean little, while relative values
are significant. Incidentally, if you run this test back 20 years instead of
15, the results for a VIX < 20 are actually significantly better than a VIX > 20
from a profit factor standpoint (and slightly better percentage-wise). This is
largely due to a huge loss that occurred because of the crash in 1987, though.
So next time someone tells you that the VIX
crossing 20 or 30 or some other number means something, you’ll know the truth.
The absolute number means little if anything — it’s the number relative to
recent levels that is significant.
Best of luck with your trading,
Rob Hanna
P.S. If you have interest in learning more about the VIX, Larry Connors has done
numerous studies, and you can find them on
www.tradingmarkets.com.
P.P.S. For those who may be looking to expand their use beyond just market
timing, my
Hanna ETF Money Flow System utilizes the VIX in generating trading
signals for spread trades.
Rob Hanna is the principal of a money
management firm located in Massachusetts. He has spent the last several years
developing and refining methods for trading in stocks across multiple time
frames. He selects stocks using both fundamental and technical criteria, and
then trades them using technical analysis techniques.