TAA Model for June, 2011
This is a new monthly feature at the MarketSci Blog.
Our Tactical Asset Allocation (TAA) model selects up to four assets from a diversified basket of asset classes on the final trading day of each month. Below is the new allocation for today’s close. Click to read more about the TAA model.
I eat my own cooking, so I’ve devoted a healthy share of my own net worth to the TAA model (read why). On the last day of each month I share my new allocation (see above) and real-time performance (see below) excluding trading frictions and taxes.
The model underperformed the benchmark in May, returning (as of yesterday’s close) -1.8% vs 0.1%, mainly due to the abysmal performance of commodities. For the upcoming month of June, I’m selling out of commodities in favor of real estate and increasing my position in U.S. Treasuries.
There hasn’t been a significant difference in performance between the model and the benchmark so far. As I’ve discussed before, I think the real strength of the model becomes apparent when equities and related asset classes go through a protracted downturn. Until then, I’m happy to keep pace with the benchmark given how diversified our holdings are.
Happy Trading,
ms
. . . . .
To stay up to date with what’s happening at the MarketSci Blog, we recommend subscribing to our RSS Feed or Email Feed.
Filed under: Tactical Asset Allocation | 14 Comments
Hi Michael,
Are you reinvesting your profits with this TAA system?
-Mark-
Hello Mark – yes (except for dividends which are reinvested month end). michael
Hi Michael,
It looks like your TAA model has more variability than the benchmark. I imagine that’s because the benchmark is 50% IEF, but I’m wondering if it’s also due to only holding 4 ETFs in your TAA portfolio. Do you have results from holding more positions?
Also, your backtests show your TAA model versus S&P500 buy/hold. How does your TAA model compare to just SPY?
Thanks,
Josh
Hello Josh – I would say it’s because the benchmark has been far less volatile since 10/2010 than the historical norm (mainly due to a lack of volatility in U.S. equities), so basically adding any additional risk assets is going to appear more volatile than the benchmark. Over the long-term I am expecting the volatility of the model to be roughly inline with the benchmark.
Not sure I understand your second question. Do you mean how would it look since real-time inception relative to SPY? Crushed of course as you would expect with any diversified approach. SPY has run at about 30% annualized since 10/2010.
michael
As always – thanks for providing updates. And to answer the above question – yes, he is reinvesting his profits.
-Jon
RE to Jon – appreciate the kind words sir. michael
Michael-
Enjoying the blog, especially the TAA stuff.
Along the lines of Joshua’s question, I was wondering if you’d considered increasing the number of assets to get more granularity as Meb Faber has done with the gtaa etf.
Also, I was wondering if you’d tested rebalancing mid-month instead of month end. I’ve seen some degradation in back tested results with a mid-month rebalance and
don’t quite understand why this would be the case.
Lastly, do you try to diversify among sectors (i.e. choose the strongest etf in each sector) as I’d imagine it would be possible to have all equity etfs in your portfolio during strong momentum periods.
Hello Steve – good questions…answers below:
1. I tried to reach a middle ground between enough asset classes to achieve some level of diversity and few enough to make the strategy useful in smaller accounts. I can’t see going below four (just too concentrated), but I think anything north of four is perfectly fine.
2. I did some testing but didn’t see a statistically significant difference in performance.
3. I don’t do sector based timing – I think individual sectors are more difficult to time use trend-following indicators than broad equity indices, but more importantly, with only a four asset portfolio, I don’t want to further concentrate the equity exposure.
Side note: in my take on TAA, it’s NOT possible to have all equity ETFs at any given time, regardless of what the broader market is doing. A fundamental tenet of TAA is to always maintain some minimal level of diversity. I think a strategy that can go “all in” on equity-like instruments (or any other asset class) isn’t really TAA, it’s market timing.
Just my $0.02.
michael
What do the results look like when commissions are included?
RE to Chris: that would of course depend on transaction costs at a given broker and the size of the account. Assuming for example, 100k acct size and a flat $5/trade, total drag on returns YTD would be about 0.1%. Slippage (so far) has been, for all intents and purposes, nil. michael
I had the same question – easiest way to do it is to multiply your cost per trade (e.g., at Fidelity it is $7 per trade) by 8 (for the maximum number of monthly trades made – includes both the sell and buy sides) and then by 12 (for 12 months in a year). Now, you won’t always make 8 trades per month – sometimes you’ll only be reallocating a portion of the portfolio. For example, this month it only took 3 trades to achieve the model portfolio. So for Fidelity, the worst-case cost for trading for the year would be $7x8x12 = $672. My rule of thumb is to try to keep trading costs, outside of the ETF expense ratio, at 1% or less to make it worthwhile to trade a strategy. To keep costs under 1% and trade this strategy using Fidelity’s commissions, you’d need a $67,200 account. However, actual CAGR does not just reduce by 1% because the cost of trading stays the same as your account grows. For example, a 15% CAGR over 10 years on a $50,000 account with no commissions is ~$202,277 while the same account, taking out $672 in commissions at the end of every year totals $188,633.
Alternatively, you could find commission-free ETFs that are similar to the ones Michael chose (such as IVV through Fidelity*) or find commission-free mutual funds that do the same thing.
*Sorry to make this sound like a marketing pitch for Fidelity, but I’m the most familiar with their products since I use them for my accounts.
Hello Jon – great comment – only one thing I would add: it’s (very) rare that the model would replace all four positions in any given month. I would have to look back for an exact number, but we’ve averaged somewhere around 4 trades in any given month (in real-time). In my own analysis I put the number an investor would need to trade with more or less no significant impact from commissions at about $50,000. michael
Michael,
I appreciate you sharing your research with TAA and sharing your monthly trades with implementation of the strategy. Did you use a testing program like amibroker, tradingblox, wealth lab, etc. to do your backtesting, custom programming, or a simple excel spreadsheet?
I’m working on doing a backtest for a similar strategy that I can implement with the tradeable instruments in one of my accounts.
Thank you in advance,
Ross