I created a Google Custom Search Gadget and added it to the World Beta toolbar. It searches only the top 50 investment blogs. Let me know what you think (and if I should add some more blogs). You can also find the search site here where you can customize the gadget or add it to your Google homepage or blog.
—-
While oil is gushing to the moon today I re-read this great PowerPoint from Edward Qian at PanAgora – Multiple Alpha Sources and Risk Management. In the presentation they have a slide that shows asset class returns broken into Fed tightening and easing periods.
I thought it would be interesting to take the five asset classes mentioned in my paper and recreate the study with data going back further to 1973. Each month I recorded the year over year change in interest rates, and then the next month returns for the asset classes. For example, if interest rates declined from 5% to 4% then that would be considered a Fed easing bucket. While any number of interest rates could be used, I simply used T-Bills (and I know T-bills don’t equal fed funds but it should be close being at the short end). Below is the chart for all months from 1973-2007. In periods of fed easing, stocks, foreign stocks, bonds and REITs all performed better which makes intuitive sense because they are all capital assets and benefit from lower interest rates. Commodities on the other hand, turn in twice the returns when interest rates are going up, which makes sense due to their correlation with inflation and unexpected inflation. A simple model would be to alter portfolio weightings based on trend in interest rates. Capital assets (stocks, bonds, REITs) should have a higher percent allocation when interest rates are going down, and ditto for commodities when interest rates are going up. If I get around to it I’ll report some portfolio performance numbers for this strategy.
The red dots are Fed Tightening periods, and the blue dots are Fed Easing periods. Click on the chart to zoom in.
It is interesting to notice that the change in environments for stocks and bonds is smaller in the level of returns than for REITs or commodities and that the volatility levels are unchanged for stocks and bonds.
I’m not sure I get your assertion that commodity prices should go up as interest rates go up. If we just think about inflation – which has caused commodity prices to go up wildly recently – all this happened while interest rates were coming down. Or am I totally off base here?
Once again I hear a lot about risk-parity portfolios – now if I only knew how to generate one.
All things considered, I would agree with you damian that I would expect gold to perform better during an easing; however, you might want to look at John Murphy’s book on Intermarket Analysis. He notes that commodities tend to move slowest in an economic cycles. However, you should note that GSCI is mostly oil. So try to think about what oil would do in a tightening and easing. In an easing that means that there is less economic activity and less need for oil (also note that industrial minerals should be off significantly in this case), vice versa for a tightening. In the case of bonds and stocks, a 50 basis point change in the interest rate is coming to change valuations very significantly compared to how the cash flows should be adjusted (well just the stock’s cash flows should be). It makes sense, but I would be interested in how much it can improve on the taa model. You would think that some of the trends are already included. Might help you get in early or out early.
I might do some work on this on Monday if Mebane hasn’t posted anything by then.
Can you explain your x-axis? It’s labelled “volatility,” but that isn’t a measure of Fed policy. Your text states you measured changes in interest rates, but those are occasionally negative, and your x-axis is only positive. Many thanks for a bit of clarification.
Damian:
The way I see it, the tightening usually occurs in response to current inflation, which helps commodity returns, while loosening combats current [disin/de]flation, which hurts commodity prices.
I could see, however, where commodity prices would crack right near the top of a tightening cycle (think oil in the early 80′s), so returns may vary during early/late parts of the cycle.
Anon: The numbers on the graph show the performance of the asset classes, NOT interest rates.
commodities is very broad and do not necessarily have correlations, hence look at base metals and crude oil for the past week. Maybe reverse correlation…..
Thanks for including my blog in your search engine, Mebane.
David
If you’re looking for some insight into the rates/commodities/stock market intermarket analysis there’s a great paper by Charles Kirkpatrick, CMT at his website on Long Wave Dow long bull/bear cycles and how tagged to interest rate trends.
He notes that the rates will sometimes trend opposite and at times together with the market, due to the fact that the stock market has growth component/corp profit and an interest rate component/alternative investment.
http://www.charleskirkpatrick.com/
FYI -
Interesting. I’ve also noticed that in Q1 2008 all the asset classes, including commodities, exhibited a far higher degree of correlation. Things seem to have returned to a more normal coupling. You can track recent asset class correlations at http://www.assetcorrelation.com
Did you ever get around to running some numbers on this strategy? Could come in handy in the years ahead. Thank you.
Did you ever get around to running some numbers on this strategy? Could come in handy in the years ahead. Thank you.