Larry,
I have another way of looking at the chart you have in the SMR. I can't get stockcharts before 1990 so I couldn't annotate it myself to be sure, but it looks like a trend line drawn from 1980 to 1995 and extended beyond is interesting. The bubble would be overvalued compared to gold, and it looks like the trend might be broken, but could it only be an over correction making stocks temporarily cheap? Will we really know until the ratio tests it's high?
The real question is why would this ratio trend? I can imagine a few reasons: the market is based on human production which can advance due to effeciency while gold can not, improved mining techniques make more gold available, currencies based on gold actually kept gold inflated and this is no longer happening. I really don't know enough about economics to know if any of this makes sense. By the way, I'm all for believing the market's overpriced. I just want to look at everything from both sides.
Howard

Howard,

Let me show you the chart again with moving averages...



A question that I'm surprised doesn't come up more often is how does one know when the primary trend has changed. For that I'll defer to the master of primary trend analysis, Richard Russell. And many times I've seen Russell use a monthly chart of something and state that when the 20 month moving average crosses the 40 month moving average and both are pointing in the same direction, it's a good indication that the primary trend has changed. As you see from the above chart, that is the case. It would take a heck of a move in the ratio to move the 20 month back above the 40 month. And it appears to me that, rather than testing the highs, the ratio is much more likely to test the 1:1 ratio area. In other words, the S&P and gold would be trading at about the same price.

As far as “why” is concerned, we could talk about a lot of fundamentals as to why the stock market would go lower and gold would go higher, but reasonable people could disagree on long-term fundamentals.

I go back to what I wrote at the beginning of 2003 about major investment themes for the future. There has been a tremendous amount of liquidity created by over-accommodating central banks in the past few years. That money is going to find a home somewhere. And it will tend to flow from overvalued assets into undervalued assets. The stock market is still coming off a tremendous 18 year bull market. Gold is just ending a 20 year bear market. So in that respect gold (and other commodities) are very under priced compared to U.S. stocks.

So that is a very short explanation of my case for higher metals and lower stocks. Obviously, many others have a different view.

Larry

Larry,
Can you do this on bigcharts? I can't go back so far on stockcharts. I'd like to look for and backtest a system of switching between stock market and commodities.
Howard

Howard,

Yes, you can do it on BigCharts.com. But if you can get a monthly chart you can do it on StockCharts.com as well. Just choose “NAV” as your “style” and choose “fill the chart” as your “predefined.” On many charts it will show you the history going back to the early 80’s.

Larry

Yep it worked, thanks,
H

Larry,
Do you know of a place I can find monthly prices on gold and the longest used stock index? I'm assuming that's the DJI.
I'd like to test this over several decades.
H

Howard, do you need the actual monthly prices or would a Dow/gold ratio chart going back to 1900 be good enough?

Larry

Well, I think I'd need the prices. The chart I need for setting up the signal but if I don't have the prices how can I calculate entry and exit. Do you think this is worth doing?
h

Howard,

I see what you want to do. I don't know if it's worth doing or not, but you never know until you test it. I know there are places where you can get monthly prices for the Dow and gold going back several decades, but off hand I can't think of where I've seen it. It will come to me sooner or later and I'll post it if someone else doesn't come up with it before I do.

In the mean time this guy has some really good historical charts...

http://www.sharelynx.net/gold/historical.php

Larry

Larry,
Thanks. I haven't checked the link yet, but yesterday since I couldn't find gold prices I tried the Yen. Found yearly data back to 1956 and monthly data from 1990. For the years before 1990, I enterpolated monthly data as a straight line between years. I know that's not accurate but I wanted to test the methodology. Using the rule you gave, 20 month SMA crossing 40 mSMA and both heading in the same direction, as the signal I found the following. I did not calculate interest in a cash account for these years so this is an underestimation of results.
$10,000
Date_____Act__________$________________yen
Dec-72__buy yen ___$10,000.00 JPY 3,030,000.00
Dec-82__buy dollar_$12,165.74 JPY 3,030,000.00
Mar-86__buy yen____$12,165.74 JPY 2,687,898.60
Jul-97__buy dollar_$22,707.60 JPY 2,687,899.00
so there would have been a doubling of value from 12/72 to 7/97, not great. But I haven't figured in results from interest bearing cash accounts in the dollar years and don't know how in the yen years. Also, I think it may depend on what we're comparing it to. I haven't figured the drawdowns so I don't know if it should be compared to the risk of a cash account or equity, etc.
Let me know if you think this is a worth while methodology.
Howard

Howard,

What you're trying to do is called moving average crossover system. I can tell you that they don't tend to work very well except in markets that are trending well.

Van Tharp in his book, "Trade Your Way to Financial Freedom" writes that Chuck LeBeau (a trading systems analyst) told him that his firm tested every possible combination of moving average crossovers and none of them were much better than random entry.

I was suggesting a 20 and 40 week moving average crossover as one way to measure the primary trend. That's very different than saying it could be used as a mechanical system for entries and exits - especially something as slow moving as 20 and 40 month simple moving averages. It's going to get you in and out pretty late.

Larry

Okay, That answers my post about ssri:paas, but I have a question. In Murphy's book, The visual invetor, he has a chapter on relative strength and rotation. If I read that right, and there no saying I did, he advocated examining ratios like the one you showed us (though for shorter time frames of course) and using them to switch to rising sectors. Why wouldn't we want to take a larger approach. Rather than limiting the captial to the us equity market, wouldn't we look for each international market and commodities. If a commodity (such as gold) is rising in strength against all international markets wouldn't we pick gold? If a market, let's say US, was rising against gold, wouldn't we then want to examine relative strength for the strongest sector and then for the strongest stock in the sector. We would probably use different time frames at different levels of analysis (monthly for primary trend, weekly for leading stock in sector).
I've learned a lot here and I'm looking forward to learning more. I'd love to know what you think.
Howard

You're absolutley right. In fact, you sort of read my mind because I've been working on that very thing for awhile now - a top down approach where we identify market direction. And then the srongest sectors (or weakest if market direction is bearish), and then the best stocks in those sectors. But I'm doing it with point and figure charts.

Let's say the COT is bearish like it is now. I'm also going to use as a confirmation the point and figure charts of the bullish percentage of the NYSE and Nasdaq for more specific timing. And then, using the point and figure charts of the bullish percentagle of the various sectors, I'm going to identify the strongest or weakest sectors when they're early in their moves. And then from there I'll look for bull or bear point and figure signals and relative strength or weakness for individual stocks.

I haven't dome much with inter-markets yet (gold vs. stocks, bonds vs. stocks, U.S. vs. international, etc.), but I think it's a smart thing to do.

Stay tuned, Howard. You're going to be hearing a lot more about this subject in coming days and weeks.

Larry