Zion National Park
On page 167 of Mebane T. Faber and Eric W. Richardson’s book, “The Ivy Portfolio: How to Invest Like the Top Endowments and Avoid Bear Markets” we read, “Our research has shown that returns are lower and volatility is higher when asset classes are below the 10-month moving average.” What investor is not interested in reducing portfolio volatility while increasing returns. And possibly more important, reducing losses. This philosophy is what is behind the ITA Risk Reduction (ITARR) model we are using with the ITA portfolios.*
The 10-month simple moving average Faber and Richardson discuss in their book is equivalent to the very familiar 200-Day Simple Moving Average. Instead of the 200-Day SMA, I use the slightly faster 195-Day Exponential Moving Average (EMA) for each asset classes used in the ITARR portfolios. Readers will find a detailed explanation of the EMA elsewhere. As a review, here are the ITARR rules. These simple rules can be implemented by any investor.
- Buy Rule: Buy the index ETF when the price of the ETF moves from below to above the 195-Day EMA of the ETF.
- Sell Rule: Sell the index ETF when the price of the ETF moves from above to below the 195-Day EMA of the ETF.
On the Sell Rule, proceeds will be moved to a money market fund or to a treasury holding such as SHY. After a Buy or Sell move, we do not pay any attention to that ETF (or other investment vehicle) until more than 30 days pass. Right now I use a 33-Day waiting period. This “neglect” rule is applied for several reasons. 1) We want to avoid the wash rule. 2) We can avoid TDAmeritrade’s high commission fee if we do not trade within a 30-day period. 3) We minimize the whipsaw problem if we neglect the ETF or index fund for at least 30 days. 4) The 33-Day waiting period rotates the portfolio review throughout the month so month-end results to not unduly influence portfolio results.
The Buy and Sell signals are included within the CWM spreadsheet, available to Platinum members. In general, we examine all ETFs held in a portfolio at the end of each review period to see where they are positioned. In addition to the ITARR model, we also use the SHY cutoff model. The SHY cutoff model is discussed in nearly every portfolio review blog.
If there are any questions related to the ITARR model, please drop them in the comments section.
* Two exceptions are the Schrodinger and Copernicus as they are passively managed.
Discover more from ITA Wealth Management
Subscribe to get the latest posts sent to your email.
Lowell, assuming I follow a very simple momentum trading strategy, such as rotating once per month into the single top-performing asset class (ETF) out of a pool of ten asset classes, based on the total returns of these respective asset classes over the past 5 months, and assuming that one of the asset classes is cash, would you see any point in applying the sell-rule in your ITARR to this strategy? One could for example move to cash when the alarm goes off and wait until the beginning of the next month before moving into the top-performing class at that point.
Thanks,
Axel
Axel,
One of the alarms I look for is an ETF or security that is under-performing SHY. The second alarm is a long standing one that must go back to the 1960s, if not earlier, and that is to sell when the price of the ETF (or security) moves from above to below its 200-Day Simple Moving Average. This is the signal Dick Fabian used in his famous newsletter, ‘The Telephone Switch Newsletter.” This signal received renewed attention when Mebane Faber and Eric Richardson published their book, “The Ivy Portfolio.” I used this signal back in the late 1970s and 1980s, only I used a 200-Day Exponential Moving Average as it is very easy to calculate with a computer.
I now use the 195-Day EMA as it puts one ahead of all the folks using either the 200-Day SMA or 200-Day EMA.
Yes, one could select 10 asset classes and remain invested in the top ten. Depending on what asset classes are selected, one might well end up in highly correlated ETFs. Cluster Weighting Momentum (CWM) analysis does in part what you are suggestion, but it also keeps one invested in low correlated ETFs.
Lowell
Lowell,
I am not sure we are talking about the same thing. I am talking about a rotation strategy that each month rotates into the ONE top-performing ETF out of ten (including cash). My question is whether you think applying your EMA195 rule daily to this strategy makes sense.
BTW, what is meant by your sell/don’t buy rule for ETFs that underperform SHY? Is this rule meant to apply on any given day, or does this apply after the underperformance has occurred for a certain period, or exactly once every 33 days?
Thanks,
Axel
Axel,
What are the nine asset classes in addition to cash?
Lowell
Lowell,
PowerShares DB Commodity Index Tracking (DBC)
iShares MSCI Emerging Markets Index (EEM)
iShares MSCI EAFE Index (EFA)
SPDR Gold Shares (GLD)
iShares Russell 1000 Index (IWB)
iShares Russell 2000 Index (IWM)
SPDR Dow Jones REIT (RWR)
iShares Barclays 20+ Year Treasury Bond (TLT)
3-month Treasury bills (Cash)
So 9 including the cash.
Axel
Axel,
My “simple” approach would be to invest (equal dollar weight) in the top 3 (at least 2) ranked assets from your list providing the momentum ranking is better than SHY – in which case the allocation would be assigned to SHY (or 3-month T-Bills). See my recent post on the Feynman Study Part 10-1 (also the original Part 5 and Part 6.2.5 Posts) for an idea how this works.
If you are using the Ranking SS but prefer to use a simpler single 5-month period simply set the ROC1 days to 150, the ROC1 rank to 100% and the ROC2 and Volatility weightings to zero and you will get your Rankings – don’t forget to add SHY to the list as a filter even if you don’t intend to invest in SHY.
This is a simple system but should work ok. The reason I would take the top 3 is to provide some level of diversification and reduce volatility. IWM and IWB will probably be highly correlated so you might end up with 67% in these 2 assets at times, but at least the 3 rd selection should be less correlated and reduce volatility/risk.
Depending on how much time you have to spend on monitoring I would probably just use a simple monthly adjustment schedule. I would suggest avoiding a knee-jerk reaction to a 1-day drop below a defined filter if checking on a daily basis – rather set a rule e.g. 5 days below the filter without crossing back above it, to exit – if you prefer a more active exit style.
You can add a moving average filter to the system, but, generally, the SHY filter kicks in faster than the MA filter.
Hope this helps.
David
Sorry, can’t edit my response – the allocation to SHY/Cash would occur if insufficient assets (<3) ranked higher than SHY.
David
David, thanks. Yes, I realize that some diversification is better. I am going to post another question in the forum, under the Portfolio Management category, that relates to this. Probably tomorrow.
The hard part about having a more fast acting risk management tool for this rotation strategy is to know what works. So far, I get this strategy and the buy/sell signals from cxoadvisory.com, a website that will appeal to anyone who is signed up with ITAWealth. They did do a study with a stop-loss rule. The rule was to move to cash whenever the asset dropped more than 8% in a single day. This only degraded performance. Diversifying across asset classes helps a little in some cases, if at least one of the classes is not strongly correlated to the others. But that doesn’t happen much or reliably.
Axel