Tactical asset allocation – june 2016 update

26 Comments

JOHN STEIN · June 1, 2016 at 6:22 am

Thanks for all the work you do , love the posts . And found the last two very
interesting ( on Unemployment rates and the 12mo MA of them ) , but a bit scary for what may be ahead . Thanks again

Peter · June 1, 2016 at 4:03 pm

Thanks, Paul. I hate to harp on this, but my numbers again show that GBM should be invested in CRED, not MBB. If you change the period to 252 days on this site http://stockcharts.com/freecharts/perf.php?MBB,CRED you can see that CRED is ahead. Could you let me know if I am doing something wrong? Thanks.

paul.novell@gmail.com · June 2, 2016 at 9:25 am
Don’t know Peter. Obviously, there is a discrepancy in the 1 year returns between the two sources. Checking with the ETF issuer site, iShares, your numbers are correct and CRED is slightly ahead.
Paul

Pat · June 1, 2016 at 4:34 pm

Hard to watch IAU dropping and going from a huge win to selling at a significant loss. A year into and down over 7% . It seems that this strategy is better on paper than when put into actual use.

paul.novell@gmail.com · June 2, 2016 at 9:09 am
Possible. All strategies underperform, often for long periods of time. Historically, trend following and timing strategies have had periods of underperformance for up to 20 years. Also, a drawdown of at least 20% should be expected at any time. That is normal.
The AGG portfolios entered IAU at the end of March with a buy price of 11.99 and exited at the end of May at 11.72, which is a loss of 2.25%.

Joe the Computer Guy · June 2, 2016 at 6:31 am

Hi Paul,
I have been reading and am trying to read through all your posts and wow such great in depth information. Thanks for sharing! I have not gone through everything yet. I also ordered the book ” What Works on Wall Street…”
I used to just leave my retirement money with an Amerivest account on TDAmeritrade (Yes, I pulled the money out and it now sits in cash). Needless to say they have under performed horribly since I started with the 18 months ago.
I have my own portion of my retirement that I manage and have done considerably better then Amerivest but no real strategy. More timing and feel than anything. I know, not good and realize I will not be able to consistently match or beat the market.
I am “retiring” next year at 52 but will not need to start drawing on these retirement funds until I am 60 (also full time in an RV). My retirement plan works with an average annual growth rate of 5%.
SOooooooo.
I was going to do the QI/TV and QI/Cons Staples Utility strategy.
I read that you only have a small portion invested in these strategies. I realize you are more conservative than I. Curious why you haven’t committed more of your portfolio to these strategies. Also, if you had a majority of your retirement in cash right now would you commit to these strategies? Or only a certain percentage? What would you do with the balance?

paul.novell@gmail.com · June 2, 2016 at 10:22 am
Hey Joe, thanks for the comment. I run a 50/50 allocation at this point in my retirement. 50% risk assets, 50% bonds. It’s the most appropriate allocation for where I am in my retirement, temperament, and where I think asset prices are. I think I have more than enough allocated to risky portfolios, for now. I do also trade a small portion of my account so that keeps my interest in risk seeking quite active and satisfied.
Successful retirement outcomes are more about minimizing drawdowns and maximizing risk adjusted returns than absolute returns.
The only advice and I can offer you is to pick your overall asset allocation first. Then pick your strategies based on risk adjusted returns and drawdowns, vs absolute returns. Then consider allocating to them over time in chunks vs all at once.

Joe the Computer Guy · June 2, 2016 at 12:46 pm
Hi Paul,
A couple follow up comments/questions.
My thinking because I have a 9+ year time frame before I will touch this money is to put a top heavy weighting towards equities (these strategies). But I need to do some more research about risk adjusted returns (do you have a post you can point me to?)
I was using 5% as an average over time and thought it was conservative considering the S&P500 average over time is considerably higher.
When you say chunks, what would you do for spacing? Each month, quarter, year over how many periods/chunks? Right now a majority is in cash so I am in a position I don’t really want to be in (although the market in general is close enough to it’s highs but that’s a topic for another time).
I feel very lucky that I have done as well as I have flying by the seat of my pants up until now. Don’t want to do it that way anymore though. Retirement is so close I can taste it.
Thanks again for your reply and time.
Joe

paul.novell@gmail.com · June 2, 2016 at 2:11 pm
I update a bunch of portfolio stats once a year. See here. Focus on the portfolios with high risk adjusted returns, measured by Sharpe and Sortino ratios. I always also look at drawdown or worst year.
I think the primary determinant of how much you should be in equities or risky portfolios is your tolerance for losses and drawdowns. The larger the equity portion the bigger drawdowns and volatility you will experience. IMO, it doesn’t matter if a certain model says you should be X% in equities based on your timeframe if you can’t stick with said model when times get tough.
As far as return expectations, based on today’s valuation, the safe bet is to set them quite low. There a bunch of good sources out there but I use Research Affiliates and their 10yr expected return tables. See here. Of course, valuations can continue to go higher as well.
When I say chunks, it’s completely up to you. There is no right answer. The historical data says it’s usually better to invest all at once but I don’ know of many people who are comfortable doing that.
Paul

Mark · June 2, 2016 at 9:33 am

This is where trailing stop losses protect me. Been doing this now for 2+ years and have only found a couple instances where I left some money on the table. The flip side is it let me lock in gains vs. giving them back.

paul.novell@gmail.com · June 2, 2016 at 10:24 am
Mark, I seem to remember you use 5% for your trailing stops, right?

Roland Olsson · June 2, 2016 at 10:54 am

Hi Paul!

Thanks for, probably the best site out there on mechanical investment strategies. I’m currently using the AGG3 strategy and even though I know that all strategies sometimes underperform for years in a row, I can’t help but wonder about these last few years. It seems the AGG3 has underperformed for quite a while now. Reading your post (https://investingforaliving.us/2015/01/09/2014-diversified-portfolio-performance-update/) it seems all years since 2010 has had a quite modest performance (except 2013, ~28%, which still underperformed the S&P500) – is this period of underperformance historically common for this strategy?

Another related question is that of the risk of a strategy becoming so common that it starts to underperform (quant strategies based solemnly on the P/B-metric etc). It would be very interesting to read an article based on this phenomena – how and why it occurs (is it caused by large funds systematically exploiting these anomalies or other investors etc).

Thanks again for a great site!

Best Regards

paul.novell@gmail.com · June 2, 2016 at 2:19 pm
Hi Roland, thanks for the comment. There’s not a lot of real world history on these portfolios. Many the ETFs and factors were not investable before 2004-2006. But looking at similar strategies where there is a longer record, yes the period of underperformance can be quite significant, a few times up to 20 years. This is one reason I always suggest using multiple strategies, and to focus on risk adjusted returns, not absolute returns. IMO, the expectation for these strategies is to outperform a broadly diversified global portfolio (not the SP500) on a risk adjusted basis (Sharpe, Sortino ratios) over a full market cycle (bull and bear). And yes, there is a risk that since these strategies are easily investable now due to rise of smart beta ETFs, that the advantage could be arbitraged away.
Paul

christian örum · June 2, 2016 at 12:00 pm

Looks like i did a misstake this month.. Maybe i looked on the 30th of may insted of 31:st.. I am invested in VTV… =) I wont change this month, will ride my error out and see what happends.
Thanks för all your work that you share!

/Christian frome Sweden.

PS: i have around 22% of my savings in the AGG3 model.

paul.novell@gmail.com · June 2, 2016 at 1:51 pm
Not a big deal. So far, it’s working out.

Mark · June 2, 2016 at 12:17 pm

Hi Paul…….yes you and I have discussed trailing stop losses before. I flucuate between 5 and 10%. 5% can bump you out of a position prematurely, but 10% can eradicate gains. I am still adjusting this “add-on” to the model. While I know back tests may show otherwise, trailing stop losses let me sleep a whole lot better at night. If I do exit a position and the fund stabilizes and still has a buy signal based on the 200 SMA, I re-enter the market. In and out cost me $7.95 x 2. So for $16 I got pretty cheap insurance, locked in gains and sleep like a baby. 🙂

paul.novell@gmail.com · June 2, 2016 at 1:31 pm
Agree. I’ve been using a 10% trailing stop for a bit now. It’s more emotion management than anything else.
Trailing stops have been a huge enhancement in performance and risk management in my stock quant strategies.

Mark · June 2, 2016 at 3:20 pm
Agree on the emotion part. I know at one point you and I had a “discussion” about this and as I recall you felt back testing didn’t support trailing stop losses. Even though I very much respect your views, I decided to implement them for my AG3 and AG6 and other portfolios. IMHO back tests are great for looking through the rear view window to test theories and offer valuable insight to investing approaches. But when looking through the windshield and the road ahead has blind spots and curves, its nice to know you have some guard rails in place to protect you from going over the edge.
Very much value your analysis and views. Maybe you could do a write-up on stop losses sometime. Hint hint 🙂

paul.novell@gmail.com · June 2, 2016 at 5:50 pm
True. And the backtests still don’t support it. But as you mention, ‘living a portfolio’ is a very different thing. Me buying-in to stop losses came through the trading part of my portfolio. I only became a consistently profitable after implementing hard and fast rules on stop losses.

Don · June 3, 2016 at 2:22 pm
When do you measure for stop losses, daily, monthly, other?

Matt K · June 3, 2016 at 7:15 pm

I’ve been running some trending value backtests on P123 and the results are significantly different depending on what month you start the back test. How do you explain or account for this? Some of the backtests I’ve run are on par with S&P 500 and some are tripling it. For example running a trending value backtest from Dec 31, 2011 to present yields 209% return vs 67% for S&P but if you start the backtest on Feb 29, 2011 it returns 69% vs 57% for S&P.

Mark · June 3, 2016 at 9:57 pm

Trailing stop losses are something I apply as soon as I take a position. If you set it at 5% and the stock drops 10% that same day after you take a position, you will sell at a 5% loss not the full 10%. However the beauty of trailing stop losses are that they adjust as the stock price increases (the whole reason we take a position, right?!) Everytime the stock increases, the 5% stop loss is readjusted based on the increased price. So if a stock increases 15% over time but has a 5% pull back, you lock in a 10% gain. (Rough math here folks for illustration purposes). Trailing stop losses can save you from big losses while allowing you to take profits off the table.

IMO set the trailing stop loss % at the max money you feel comfortable losing on a position. Hint…the lower you set it the higher the liklihood you will get bumped from a position. Anything less than 5% produces lots of closed positions based on market noise. Somewhere between 5 and 10% is where I personally try to be.

Joe the Computer Guy · June 6, 2016 at 7:15 am

Not sure if I should post this here or in a more appropriate topic but here goes anyway.
It relates to some comments I’ve read on the blog previously.
I created a complex yet simple spreadsheet to calculate my retirement and when I can retire. It takes income, expenses, inflation, savings and a return rate for savings. It calculates this all out to my life expectancy. I realize this is not perfect but I haven’t come up with a better way.
I am using the average inflation rate from the bls website. My question is around the return rate. I am using the S&P average. I realize this will be higher and lower each year. I also read (I think it was your site) how the first year return rate impacts the success of a retirement plan.
I did a cross check against SWR of 4%. Mine is a little higher than that.
Wondering what your thoughts are with the use of this logic (average return rate from S&P) in my calculations? Maybe a better rate to use?

paul.novell@gmail.com · June 8, 2016 at 5:12 pm
Hey Joe, sorry for the late reply. I was traveling for a couple of days. I’ll give you some simple answers to your questions but they can get a lot more complicated and vary quite a bit depending on your personal situation and temperament as an investor.
As to return rate for equities. Most likely the average historical return is not the best one to use going forward from this point. Valuations for US equities are elevated and returns from elevated valuations tend to be lower. Also, since your portfolio won’t be 100% equites the returns of other assets matter as well, in particular bonds. And as you alluded to what really matters is the returns over your first 10 years of retirement. Negative returns in the early years can decimate retirement portfolios. Do a search on SWR on the blog and you’ll find lots of posts discussing this topic. A 4% SWR has worked in the past in all kinds of market environments. I wouldn’t use any higher than that plus I would combine that with a flexible withdrawal plan to maximize the chances of your portfolio lasting.

Mark · June 6, 2016 at 2:44 pm

Sorry about the double post. Sometimes the system takes the comment and immediately shows up. But lots of times it does nothing after I hit Publish, making me think it didn’t take. I will try to be more patient.

paul.novell@gmail.com · June 7, 2016 at 8:55 am
No problem Mark. Easy to edit.