One of the hottest topics over the past couple of weeks has been the
extremely narrow range into which many indexes have curled.
There are any number of ways to measure this, but the most popular is by
using volatility bands, such as Bollinger Bands (BBands). This
simply plots an upper and lower band around price, using a 2% deviation
from the 20-day moving average.
The chart below is an example using the S&P 500, along with the spread
between the two bands. From this, we can see just how much those
bands have squeezed together lately.
The textbooks say that when we see such a coil, then the market should
be in store for a large move one way or the other, but typically in the
direction of the next break in price.
Now that the S&P is close to closing above its upper Band, let's go back
to 1928 and look for all other times that the index has seen its BBands
tighten so much that they are less than 2.5% away from each other.
The table below shows the returns going forward when the S&P closes
above its upper Band.
|
S&P 500 Performance After A
Breakout From Tight BBands |
|
(76
trades) |
1 Day
Later |
5 Days
Later |
10 Days
Later |
1 Month
Later |
3 Months
Later |
| Avg Return |
0.3% |
0.3% |
0.5% |
0.8% |
2.4% |
| % Positive |
70% |
64% |
64% |
62% |
70% |
| Avg Win / Avg Loss |
1.9 |
1.0 |
1.0 |
1.1 |
1.3 |
| Any random
time... |
| Avg Return |
0.0% |
0.1% |
0.1% |
0.5% |
1.5% |
| % Positive |
54% |
55% |
56% |
58% |
61% |
The returns were OK, especially the day after. Across all time
frames, the S&P actually out-performed any random time, so maybe there
is something of an edge to trading with an upside breakout.
Let's further refine our trade to those times when the index is in an
uptrend, defined as a rising 200-day moving average:
|
S&P 500 Performance After A
Breakout From Tight BBands
When 200-day Average Is Rising |
|
(57 trades) |
1 Day
Later |
5 Days
Later |
10 Days
Later |
1 Month
Later |
3 Months
Later |
| Avg Return |
0.3% |
0.5% |
0.6% |
0.9% |
2.2% |
| % Positive |
70% |
72% |
67% |
65% |
70% |
| Avg Win / Avg Loss |
1.7 |
1.2 |
1.2 |
1.1 |
1.3 |
| Any random
time... |
| Avg Return |
0.0% |
0.2% |
0.4% |
0.8% |
2.2% |
| % Positive |
53% |
57% |
59% |
61% |
66% |
Again, the short-term results were promising, with the index
significantly out-performing random between 1 and 5 days out.
After that, the edge dissipates a bit.
Let's add another step and look at those times when it not only broke
out during an uptrend, but also when the S&P had hit a one-year high
sometime during the past month:
|
S&P 500 Performance After A
Breakout From Tight BBands
When 200-day Average Is Rising
And Near A New High |
|
(33 trades) |
1 Day
Later |
5 Days
Later |
10 Days
Later |
1 Month
Later |
3 Months
Later |
| Avg Return |
0.2% |
0.3% |
0.6% |
0.6% |
2.5% |
| % Positive |
70% |
67% |
64% |
58% |
76% |
| Avg Win / Avg Loss |
1.1 |
1.0 |
1.4 |
1.2 |
1.5 |
| Any random
time... |
| Avg Return |
0.0% |
0.2% |
0.4% |
0.7% |
2.3% |
| % Positive |
53% |
57% |
58% |
60% |
67% |
Again in the short-term, the S&P did pretty well on the upside.
After a couple of weeks, though, it really wasn't any better than if it
hadn't hit a new high (a bit worse, in fact).
Let's add one last condition, and that's some seasonality:
|
S&P 500 Performance After A
Breakout From Tight BBands
When 200-day
Average Is Rising And Near A New High
During December |
|
(4 trades) |
1 Day
Later |
5 Days
Later |
10 Days
Later |
1 Month
Later |
3 Months
Later |
| Avg Return |
0.3% |
-0.3% |
0.6% |
-0.7% |
0.2% |
| % Positive |
100% |
25% |
75% |
50% |
25% |
| Avg Win / Avg Loss |
N/A |
1.7 |
0.8 |
0.5 |
3.6 |
| Any random
time... |
| Avg Return |
0.1% |
0.3% |
0.9% |
1.9% |
4.5% |
| % Positive |
51% |
60% |
65% |
72% |
76% |
We only have four historical trades here, so the sample size is
exceedingly small, but these kinds of breakouts during December weren't
exactly raging buy signals. After a week, in fact, three of the
four instances showed a negative return. Same for three months
later.
Overall, there does seem to be a tradeable edge to the upside over the
very short-term when the S&P breaks out of a very tight range.
That includes times when it is in an uptrend, and has traded at or near
a one-year high (though those two factors weren't major improvements to
the strategy).
The fact that it's December doesn't help any either, and in fact may
actually be a hindrance in the intermediate-term since markets often
experience a downdraft during the early-to-middle part of most Januarys.
Jason Goepfert
Founder, Sundial Capital Research, Inc.
December
21, 2009
