OK, it's that time again where the market hits a moving average point and everyone becomes a moving average expert. It's easy, you can point to a chart during a cherry picked timeframe on a cherry picked index and make it look like it's a flawless entry point. Sorry if this has a bit of a negative, smart assy tone, but to me the only thing that's flawlessly predictable about index SMA encounters. There will be countless 'experts' coming out of the woodwork telling us about how wondering moving averages are for market timing. The last time this happened, I did a little research of my own regarding whether or not the SMA alone was a good indicator for market timing. The results can be found here. I'm updating my SMA test results with a twist, I'm going to pit the simple moving average against the exponential moving average.
OK, enough of my jibber jabber. This time we're going to run the same system, but instead of using the simple moving average, we'll substitute with an exponential moving average.
Just like last time, the data is from Yahoo. The system is done with RightEdge and I posted the trading system on the RightEdge web site for download here.
Here are the inputs:
IXIC = Nasdaq Composite. Data available was from 1972 to present
GSPC = S&P 500. Data available was from 1960 to present.
DJIA = Dow Jones Industrial Average. Data available was from 1960 to present.
I've also rerun the SMA strategy with the most up to date data and so that we can compare them side to side. I've also included buy and hold results, so we can compare three ways. The starting capital is $100,000.
The results are in:
| Symbol |
Moving Avg |
APR |
Net Profit |
Buy & Hold Net Profit |
B&H APR |
SMA Net Profit |
SMA APR |
APR Diff |
Net Profit Diff ($) |
| DJI |
50 |
5.46% |
$1,278,327 |
$1,152,863 |
5.08% |
$861,335 |
4.68% |
0.78% |
$416,992 |
| DJI |
100 |
5.65% |
$1,402,050 |
$1,152,863 |
5.08% |
$861,583 |
4.68% |
0.97% |
$540,467 |
| DJI |
150 |
5.33% |
$1,192,812 |
$1,152,863 |
5.08% |
$1,163,951 |
5.26% |
0.07% |
$28,861 |
| DJI |
200 |
5.55% |
$1,332,706 |
$1,152,863 |
5.08% |
$1,295,105 |
5.47% |
0.08% |
$37,601 |
| DJI |
250 |
5.00% |
$1,012,387 |
$1,152,863 |
5.08% |
$1,229,113 |
5.37% |
-0.37% |
$(216,726) |
|
|
|
|
|
|
|
|
|
|
| GSPC |
50 |
6.19% |
$1,838,655 |
$1,289,268 |
5.48% |
$1,260,482 |
5.42% |
0.77% |
$578,173 |
| GSPC |
100 |
5.82% |
$1,527,160 |
$1,289,268 |
5.48% |
$1,072,032 |
5.10% |
0.72% |
$455,128 |
| GSPC |
150 |
6.61% |
$2,247,540 |
$1,289,268 |
5.48% |
$1,991,021 |
6.34% |
0.27% |
$256,519 |
| GSPC |
200 |
7.05% |
$2,786,651 |
$1,289,268 |
5.48% |
$2,215,377 |
6.56% |
0.49% |
$571,274 |
| GSPC |
250 |
6.68% |
$2,326,211 |
$1,289,268 |
5.48% |
$2,638,692 |
6.92% |
-0.24% |
$(312,481) |
|
|
|
|
|
|
|
|
|
|
| IXIC |
50 |
11.03% |
$5,435,751 |
$1,749,420 |
7.90% |
$6,964,217 |
11.74% |
-0.71% |
$(1,528,466) |
| IXIC |
100 |
10.52% |
$4,531,852 |
$1,749,420 |
7.90% |
$3,404,356 |
9.72% |
0.80% |
$1,127,496 |
| IXIC |
150 |
9.82% |
$3,532,660 |
$1,749,420 |
7.90% |
$3,198,891 |
9.54% |
0.28% |
$333,769 |
| IXIC |
200 |
9.37% |
$3,009,501 |
$1,749,420 |
7.90% |
$2,961,097 |
9.33% |
0.04% |
$48,404 |
| IXIC |
250 |
8.80% |
$2,444,567 |
$1,749,420 |
7.90% |
$2,390,753 |
8.74% |
0.06% |
$53,814 |
There are some notable omissions. There is time spent out of the market, so loss of dividends is not factored in. Second, transaction costs. I didn't put the # of trades per system for lack of room, but the trades per index averaged about 10 per year. Third, taxable events. Buying and selling like this could trigger multiple taxable events. And for U.S. investors, the short term gain rate is higher than the long term rate.
The results for EMA seem a little bit better than SMA, except for the Nasdaq where they seem about the same. We're talking about less than 1% APR on either side (whether EMA did better or worse than SMA). I'm not sure this is something to shout from the rooftops. To me, this seemingly slight edge is within the range of statistically insignificant.
But let's say a 1% additional return was significant enough and sustainable, on average, the EMA based trading systems had more trades than the SMA trading system. So trading costs and taxation should be a large concern.
An interesting development here is that almost all of the moving average systems have now done better than buy and hold since I last ran the test. It just so happens that the moving average was a great indicator to get out and saved you a ton of drawdown this go around. It might be fun to run this again in another couple of years to see how whipsaws and potentially a new bull market play into the results.
Subscribe