Sunday, April 20, 2008

Analysis of the 20/50 weekly moving average Indicator

The article I am about to write is about forecasting/trend spotting using the indicator, after of course, the preliminary analysis to show that it's predictive long term ability is excellent (nearly flawless).

This particular indicator says:

  • Long-term Buy signal if: 20-weekly moving average on the Index goes up OVER the 50-weekly moving average by 1% from UNDERNEATH.
  • Long-term Sell signal if: 20-weekly moving average on the index goes down UNDER the 50-weekly moving average by 1% from OVERHEAD.

My skepticism was initially great against filter created of existing/past prices but that quickly went away after I analyzed that this filter hasn't generated a single false signal in the last 40 years. Study carefully the chart below for the SPX index from 1990 to Present time:

The white circles with Buy/Sell signals tracked the point at which the indicator generated a long term buy/sell signal. As you can see it is hardly disappointing. If you had followed this simple strategy, over the past 30 years, you would have outperformed the S&P500 index by 400%. ( Or more if you had choices of different asset class other than SPX. Like Sell the SPX in 2000 and buy real estate in FL instead :D )

Anyway, time for prediction. A question may arise at this point: Why, am I, the fundamentalist and earnings calculator, use techincal indicator as a prediction tool?

Prediction doesn't have anything to do with actual economic reality or earnings. If people started buying again for whatever reason, thinking that not-paying debt is NO PROBLEM. That they can party TWO more years with negative or ever decreasing earnings from here on, then trend-analysis is your friend. And I try pick the most reliable friend, the 20-50 Weekly moving average indicator.

  1. Look at the RED box and you understand what I mean. The 20-weekly moving average, contrary to your instinct, cannot catch up the 50-weekly from underneath by having a couple of days of massive rally.
  2. Quite the contrary, the way to get there is meandering with violent whipsaw of 10% up and 10% down for nearly half a year or more, with the final cycle being the big "DOWN", and the Buy signal would have been generated from the indicator.
  3. This is a feature of any lagging filter. The 50-week average is lagging the 20-week. You first need to stabilize the market at some level, say 1300-1400. And then violent whipsaws - typically down - enough to force the 50-week "catches" the 20-week.

To hell with realities right? Well, I haven't degenerated quite that badly.

At this point there is a glaring inconsistencies in the RED box comparison, in the sense that we are WAY up to the previous RED box in both price but NOT earnings. The fuel is spent back in 2000-2001 for earnings. So where will earnings need to come from? Selling to foreign country will be limited to infrastructures, and if attempted will be simply grabbing LOW HANGING fruit. Since I don't have the answer to this question, I will leave it to your imagination/analysis.

Time will prove that either I am right that the Long-term trend is down, or for some absurd reason it's not about earnings anymore as much as it is pride of ownership "society" when buying stocks and the Long-term trend is actually up.

But what is without a doubt from all the realm of possibilities that I have discussed here, that in the short run (6-months),
  • Buying short term call and put would be disastrous.
  • Selling long term puts is risky at best.
  • Opening a stock position long is still unnecessary given you aren't giving back a whole lot if the indicator is somehow false at this point. And by implication, opening a short stock position other than current positions.

Beware of big 10-15% whipsaws in the coming weeks.

(Credit to Denninger at TickerForum who alerted me first of this indicator, which he said he got from a "grizzled" old trader at CBOE way back.)


Penn said...

After reading your article, I cannot follow your chain of thought.

Based on the indicator, why do you feel buying short term puts is going to be a disaster?

MTGSPY said...

high likelihood it won't payoff. And also think about the implied volatility at market "bottoms" (not saying this is at all) - IV is typically highest there.

Therefore there's usually limited amount of money to be made from out of the money short term options given the entry cost is high enough to curtail leverage.

The 10-bagger of yesteryear is discounted out of the market.

Penn said...

What I am trying get at is based on the 20/50 indicator alone, how do to reach the conclusion that short term puts are unlikely to pay off?

If you made that statement based on some other factors, that is fine. However, if that conclusion was reached solely on the basis of the 20/50 indicator, I will appreciate it if you can shed more light on it

A said...

Does this work for the 87 crash? I think that is only instance where this 20/50 crossover would fail. Unfortunately my broker's charts doesn't go back that far to test it.

A said...

How do you define long term for put/call options? 6 to 9 months out or more like out to Dec 2009?

MTGSPY said...

sorry didnt get back to u earlier,

theory suggests a meandering market like what's in the red box won't reward a 1-month or 2-month (my definition) of short term option, be it call or puts.

It's all game of probability.

The puts probability of crash has for now imputed to the point that if goes down 50-60 pts on SPX the premium faces a "drag".

The calls aside from obviously at the end of a BIG rally of 150 pts is running out of space and it seems like can only lose money. Also, from here on, the more it goes up, severe imputation of probability going further up, just as in the case of the puts.

Put it this way, either choice

a) buying very OTM options 29 cents with risk it turns to 35c or 0c
b) buying close to ATM $3.15 option in the hope to turn it into $25c or $5.20 in the next month.

IS NOT attractive. Those numbers are only for your consideration. I factored in the implied volatility, the "imputation" of probability (aka mean reversion) into the pricing of those scenarios.

Hope that helps.

doncleland said...

I have always valued your analysis and insight.

Good Lucky with your new site.


MTGSPY said...

Thanks Quads (don!) :D

Good to see you here.

wolfman said...

I have an excellent market timing indicator. It's very simple too. I use the 10 day Exponential moving average and a moving average envelope over the DJIA. The envelope is a 20 EMA and the two lines are 2.5% above and below it. When the "fast" 10 day EMA crosses above the upper envelope line, the market is in a rally and go long, when the 10 crosses under the bottom envelope line go short, because it's in a correction. Most likely bear barket/recession. I checked this back to 1980 on Stockcharts . com (for free) and it avoided EVERY bear market, sidesteped the 87 crash, got out long before the 2008 disaster, and hopped on for every bull market still early in the move. It works and I love it.

goodlux said...

looks like it is crossing under again...time to dump stocks?

Stock Trend Investing said...

Like Wolfman, I use another simple market timing indicator that is doing very well for me. It is based on the patterns in the monthly closing price of a stock market index. It is not only working for the S&P 500 but also for other US and foreign market indices. And it allows me to review just once per month if I buy, sell or hold.