Dr. Craig Israelsen's presentation, "Inflation And Investing Since 1970," received 4.8-stars from attendees of the live session and A4A members can skim it or replay the entire for CE credit on demand 24/7.
Dr. Israelsen answers questions from attendees and me incisively below. Please keep the questions coming. This can be an intelligent conversation and allows advisors to publicly state their names and their views. And if you have content to share on your website and you are an advisor -- with original ideas -- please include links to youre website. It will be good for SEO and help you spread the word about your skills and knowledge.
Mohamed El-Erian recently said, "We have gone through massive roller coasters the last two years, which suggests that you have to ask yourself, ‘Do I also need to be tactical as well as strategic in the way I invest?’ You need the ability to move quickly when you get conviction as to where this is going to end up.” How would you respond to El-Erian?
It really depends what is meant by the term “tactical”. Does it mean moving a portfolio entirely to cash? Or entirely to one particular asset class? That would represent “unbounded tactical”—of which I am not a fan. A version of tactical that makes some sense to me is subtle tactical moves WITHIN a strategic approach, such as the 7Twelve model. In other words, all 12 asset classes are always in play, but an internal “tactical” move may be to move from an intermediate term bond fund to a short-term bond fund based on your outlook on interest rates. Or, to move from a blend large cap equity fund to a value large cap equity fund. Or to move from a hedged to an unhedged international bond fund. These types of “tactical” moves represent actions that trim the sails, rather than entirely moving the sail.
Please do the correlation seminar you mentioned. And, if the correlations are changing so much, what is the logic of a go-forward 7/Twelve model?
Thanks, will do. Correlations are always changing and always will change. The logic of diversification is always the same: inasmuch as we don’t know the future, we diversify.
When is the best time to rebalance? Any time to avoid?
Spring tends to be best time of year (March or April). But, the differentials in performance are not huge based on the month you choose to rebalance. The key is to not rebalance too often. Quarterly or annual is often enough (similar results either way).
It could be argued you are picking 15 years because the Vanguard 60/40 has outperformed over the past five and 10 years.
Longer time frames are generally preferred when comparing performance. But you are correct, a 60/40 model has outperformed the 7Twelve model over the past 5 and 10 year periods.
Have you considered whether a value tilt would enhance or change this analysis?
Yes, the “live” versions of the 7Twelve model that use actively managed funds and/or passively managed ETFs use several funds that have a value tilt—particularly in the small cap US equity category.
Regarding Slide #2:7 is inflation US based for all these seven asset classes?
Yes, I’m using the CPI from the Bureau of Labor Statistics – so it is US-based inflation.
Have you run these numbers over the last 30 years rather than 45?
Good question. Here are the results if 1970-1979 is removed, giving us a 36-year period. The average REAL return for commodities during the 18-years of below median CPI (which was 2.82% with the 1970’s removed) was -6.19%. During the 18-years of high inflation commodities had an average REAL return of 11.80%--the highest of any of the asset classes. By comparison, large cap US stock was 8.13%.
If we only look at the last 30 years (1986-2015) as you suggest, the median CPI becomes 2.68%. During the 15-years of low inflation the average REAL return of commodities was -9.63%. During the 15-years of higher inflation, the average REAL return of commodities was 16.89%--the next closest was large cap US stock at 8.35% and real estate at 8.01%.
What is the real-time performance (not back tested) for the portfolio since 2008?
S&P calculates the daily performance of the 7Twelve model (http://www.customindices.spindices.com/indices/custom-indices/lunt-capital-7twelve-moderate-index).
The ETF-based version of the 7Twelve model that I have built and monitor has produced the following returns:
If the average inflation number changes (‘70s drops off), how does it change the results?
See answer above.
Slide 44: I expected better results with commodities. What’s up with that?
Interesting question. The last two years (2014 and 2015) were really tough years for commodities. If we look at a 44-year period from 1970-2013 here are the results.
A 7-asset model that included commodities: 10.33% annualized return
A 6-asset model that excluded commodities: 10.01% annualized return
Slide 46: When you say 79% of the time, the two assets move in the same direction, over what period? Is it 79% of the 30-day periods over the past 46 years? What is the period used to measure the correlations. And is it more accurate to use short periods or longer ones to measure the coefficient over the 46 years?
Over the entire 46-year period based on annual returns. In other words, 79% of the annual returns of US small stock moved in the same direction as the annual returns of US large cap stock. Good question about the length of the time period. Morningstar generally measures correlation over rolling 36-month periods.
I don’t have monthly performance data for US small stock back to 1970, but I do have annual returns back to 1970 and monthly data back to 1979.
So, I did an experiment for you: the correlation of monthly returns between large US stock and small US stock from 1979 to 2015 was 0.833 and if using annual returns between 1979-2015 it was 0.815.
Very similar results. If correlation over a long time period (such as 46 years or 40 years…) is being measured then using annual returns is fine. However, if you are wanting to look at correlations over the past 3 years, it would be important to use monthly data.
You said correlation coefficient was one of two important portfolio statistics. Did you also say that another key statistic reflects the amount of influence one asset class has over the other? Please explain that a bit more detail.
The other important statistical measure is beta. It is the measure that tells us about the magnitude of correlation. Beta is a measure of the volatility of a fund relative to a benchmark index, such as the S&P 500. A beta of 1.00 indicates that a fund has volatility equivalent to its market benchmark index (e.g. S&P 500, Russell 2000, etc.). A beta coefficient of 1.10 indicates a fund’s monthly returns are 10% more volatile than its benchmark whereas a beta of .85 indicates a fund is 15% less volatile than its benchmark. A very low beta may not indicate that a fund has low volatility, it may simply be due to a very low correlation between that fund and it’s benchmark index (as measured by the correlation coefficient). Therefore, beta is a more reliable measure as the correlation is higher.
Even if an advisor practices this way, people won’t stick with it in times of distress. Explain what you said about needing to remind investors of these lessons when they are not needed.
Sadly, you are correct in many cases. But some clients are teachable and will learn to become less reactive and more proactive as investors. Thus, the education process needs to be in high gear all the time, even when markets and asset classes are doing well. Very simply, a diversified portfolio will always have some asset classes that are “under-performing” the best performing asset class. This is an obvious statement—but investors often lose sight of the obvious when fear and greed get in the way. Successful investors don’t give up on asset classes when they are struggling—in fact they put more money into them (which is what we call rebalancing!). The other reality is that investors can always put more money in their accounts if they want to see the balance go up. The portfolio itself shouldn’t have to do all the heavy lifting. Annual or monthly additions to the portfolio is their part of the agreement. Investors who invested during the decline of 2008, for example, were well compensated in 2009 and beyond. I suppose I would summarize it this way: when a client’s portfolio is doing well I would caution them that a 22% annual return is not “normal”, so don’t set bizarre expectations. Likewise, when a portfolio is struggling through a period of low or negative returns, I would similarly remind them that this is not normal either. The portfolio’s performance will always regress toward the mean. In most cases, a diversified equity and fixed income portfolio has a long-term mean return of somewhere between 7% to 9%. The short-term performance of a portfolio is often distant from its longer-term performance. Sort of like human beings that are having a cruddy day and acting like it—let’s cut ‘em some slack. Normally they are a really decent person—and that’s how we choose to see them and interact with them.