IBD Market pulse comments by xuyuzhu

VIEWS: 6 PAGES: 2

									At last meeting VAS mentioned an analysis started by Kevin Campbell in the SIPSIG Yahoo group that
looked at the effectiveness of the IBD Big Picture Market Pulse as a market timing measure. He noted
with concern that following the past six months of IBD market signals, you would have lost money. I
decided to expand on his look back using data back to 5/1/2008.

Here is a summary.

The Data
Since some of the "Big Picture" wording has changed over time, I interpreted all IBD signals as one of:
"Market in Correction," "Uptrend Under Pressure," or "Confirmed Uptrend." (see related Excel file
posted under files.)

Since the Current Outlook is published the prior evening, I am showing the Current Outlook as applicable
for the next day. E.g. when you look at the Big Picture dated 12/20/2011, the "In Confirmed uptrend" is
shown as the signal (Outlook) for 12/21/2011. That is, the first opportunity to trade the signal is on the
open of 12/21.

Signals
Similar to the other analysis, I simulated trades as follows.

o - Buy the SPY on the open of a "Confirmed Uptrend," Sell when signals downgrades to "Uptrend Under
Pressure" or "Market in Correction."
o - or, Short the SPY on the open of "Market in Correction" and Cover when signals improve to either
"Uptrend Under Pressure" or "Market in Correction." Note: I simulated going short on the SPY instead of
using an inverse ETF.


The Results

I've attached an Excel file that lists the signals and includes a bitmap picture of the trades. You should be
able to expand the picture by dragging a corner. The lavender shaded areas represent "Market in
Correction" while the pink areas are "Uptrend Under Pressure." Buys are shown as solid green arrows,
and Covers are hollow green arrows. Similarly, Sells are solid red arrows and Shorts are hollow red
arrows.

For the period of 5/1/2008 to 12/31/2011, there were 59 signals (37 long trades and 22 short trades).
Following the above rules would have actually made money over the entire period, though during 2011
the results were negative.

Some statistics (assuming beginning with a $100,000 investment):.

Buy and Hold of the SPY (purchased at 138.13 on 5/1/2008 and sold at 12/30/2011 at 123.93)- your
$100,000 would have dropped to $89,700 or down a little over 10%.


                                     All               Long Only        Short Only
Number of Trades                 59                   37                 22
Percent Winners                  44%                  49%               36%
Profit(loss)                   +34,062              +25,929          + 8,133
Maximum drawdown              -19.2%               -17.9%           -27.0%

Buy and Hold loss             (10,280)
B&H Max Drawdown               -45.3%

So while the winning percentages are not very impressive, perhaps the most important factor is that it
got you out of the market with a less terrifying drawdown, and with a small profit compared to Buy and
Hold.

The bitmap also includes a histogram showing where you would have experienced profits and losses.
You can review the largest gains in losses in the Excel file, but you should note how important it is to
follow the "Market In Correction" signals. While the market does not always go into a free fall - if you
had delayed, even a day or two, for the large drops such as occurred in Sept 2008 or Aug 2011, your
results would have been much, much worse.

A couple of points in conclusion

Subjectively, it appears to me that IBD signals seem to work better in clearer uptrends or downtrends.

The signals do not seem to work well in sideways/ trading range markets.

The signals sometimes seem to be late in switching to an uptrend - notably a few times in 2011.

In reality, you will be trading individual stocks as opposed to the index, and many IBD followers trade
long-only. However, IBD signals are important to follow - if nothing else as a reality check on when to
change your market perspective. It is also important to note that if you delayed acting on the "Market
Correction" signals, even for a day or two, would have resulted in a few huge losses for several of the
major downturns -- so take those "Market Correction" signals seriously.


Follow-up

Does anyone have experience with other market timing signals/indicators that catch these turns better?
I would certainly be interested in hearing about them and would be happy to backtest ideas if I can.

Also, I would be interested if anyone knows of any available history on IBD follow-thru days or
distribution days? They may be interesting to research further.

								
To top