Monday, September 30, 2013

Asset Allocation

For some horrible reason, I can't find the article that I was reading that was talking about asset allocation. But basically, it was saying that your choice of asset allocation is much more important than the actual individual securities of each asset class. The allocation part makes up a much larger part of your returns compared to the individual security selection.

Now, that is actually a nice fresh look at things again. While everyone is always focused on picking the best securities within an asset, people just look broadly within 2 assets classes, equities and bonds.

I am definitely am a firm believer that the subtle differences within the types of asset classes really can make a large difference to the portfolio performance, especially the OW-ing and UW-ing of asset classes.

What brought me to suddenly think about this was reading this post on Investment Moats. It's actually a link from Mebane Faber about popular asset allocation strategies and historical backtesting. The link that he gave was this one, which includes a lot of data, as well as the portfolio constituents. However, if you look at the newer posts on Faber's website, he has actually included 3 more portfolios, have a look here. It is also rather recent at 24th September too.


Now, of course knowledge is out there for us to read, review, understand, dissect, and ultimately to be made use of in the best possible way for ourselves.

Looking at the list of asset allocation strategies, I'm narrowing in on the Sortino Ratio, rather than the good old-fashioned Sharpe Ratio. The Sortino Ratio is a far superior measure, because it only uses the downside volatility, as compared to the Sharpe Ratio. This means that more attention is paid to the drawdowns, rather than the overall volatility of the portfolio. I LOVE it, because I hate looking at large downside risks. The inclusions of max drawdowns are also lovely, because I do like to know how much volatility should be coming my way.

I must say that the CAGR on all the portfolios are definitely delightful to me, all over 9% in nominal terms, except for the permanent portfolio.

Looking at just CAGR allows us to eliminate half of the portfolios. Looking at the Sortino Ratio also helps us eliminate a lot of them too. InvestmentMoats does rightly point out that the Risk Parity Portfolio has a large 70% allocation to bonds, which have been doing significantly well since 1980, which may well account for the fact that is has done so well to date, since interest rates are at historical lows and are only sure to rise.

Given that, I would definitely rule this portfolio out, knowing how that bonds will definitely not be enjoying its same run that it has had over the last few decades. However, if the world is at interest rates high and heading downhill again, I think I'd definitely jump on that bandwagon.

So, what's left actually? The Permanent Portfolio sounds tempting, with a fantastic Sortino Ratio only bettered by the Risk Parity Allocation, but it only has a CAGR of 8.88%, which is the lowest of all the portfolios! Of course, it also has the lowest volatility as well.

The traditional 60/40 looks fantastic to me, but given that the 7/12 portfolio is also a new addition to their asset allocation strategies, let's look at this a bit closer. I was just reading of the iShares blog about the 60/40 allocation, and I actually have been wondering a lot about its magical numbering and powers. This post effectively shed a lot of light on the matter, showing how the 60/40 is actually very close to being the simplest, yet most optimal portfolio. It notes though that "It is particularly interesting to note that the varied optimal allocations are primarily driven by variations in the bond component rather than the equity component." Their recommendation is this: "the key point is that investors going forward need to have a view of how bonds will perform over the next several years."

Basically, if you were too lazy to read it, let me summarize. How bonds perform in general should be the main driver of the amount of bond allocation. But, 60/40 is a great place to start at, with 43% being the historical optimal point.

7/12 is a lot more desirable given that it is a lot more broad based and diversified globally. The idea of 60/40 just being in the US market does not appeal to me at all. Given the fact that I am very bearish about the US market, their general economy, as well as their currency, I would not bet my portfolio on the US market.

Now that leaves us with El-Erian, Bernstein, Arnott and the GMI (All Market) with over 10% CAGR. Just looking at the Sortino Ratios, El-Erian and Berstein are definitely out, clocking in the lowest 2 SR's at 0.55. Berstein's strategy of plain vanilla asset allocation also doesn't appear very diversified to me, which probably shows this underperformance. El-Erian gets plus points for included an 8% allocation for special situations, which basically means gambling on gut feelings. However, I think his very low allocation to just 19% bonds, as well as allocation into private equity is not something feasible for me. Again, I love the idea of a gambling allocation, haha.

Now, Arnott and the GMI. Now, if you don't know who is Rob Arnott, that's all right. I didn't know who he was either. But, now I do. He's a really smart guy, very practical, very academic and very much emotionally detached from chasing a rally. His best quote he made in a video interview was about how to know when to sell and move on into other asset allocations. "You don't want to be the last few people picking up the nickels and pennies in front of the steamroller coming". He's also the ONLY fund manager in PIMCO, who actually doesn't work for PIMCO. Now, that says something too. He runs the All-Asset portfolio there, which honestly, I'd love to dump some money in there, haha. Also, he's the guy bumping heads with Bogle from Vanguard by coming up with another index to challenge the traditional indexing that Vanguard made so famous. He also recently reconfirms that he thinks US stocks are at least due for a correction some time soon. Now, this GMI, who came up with that? Well, one of my weekly reads from the Capital Speculator! Actually, I wanted to list James Picerno and his website on the sidebar, but his posts usually contains a fair bit of jargon and technical stuff, which even I can find hard to digest if I'm not paying attention. However, his book as well his GMI has a very impressive track record. He would be the first to admit that it isn't the perfect portfolio, but considering it outperforms 70% of other portfolios while doing it passively, I think you can stick to his strategy and still be able to brag about beating the market at the end of the day.

Arnott has got 10.13% while the GMI has got 10.14%. Arnott has a slightly higher Sharpe Ratio (0.59 vs 0.56), but also a slightly lower Sortino Ratio (0.73 vs 0.76). Their drawdowns are very close, with Arnott losing with -32.50% compared to -29.11% from the GMI.

Arnott's portfolio is very straight forward with equal weighting, making it for simple investing. I like it because it outperforms the 60/40 and the S&P from this study here. My gripe is that it is not easy to invest in TIPS, as well as the heavy allocation of 60% bonds here.

GMI's portfolio is somewhat similar as well, especially if you compare against its equal weighted portfolio. While Arnott has 10 bits divided to have 10% weight each, GMI has 13 bits. 1 part cash, 1 part commodity, 4 parts equity (US, developed, EM and REITS), 7 parts bonds (TIPS, foreign TIPS, US bonds, foreign developed bonds, EM bonds, HY bonds and foreign corporate bonds). That makes it to roughly 53% bonds, 31% equity and 8% commodities and 8% cash. Again, like Arnott's portfolio, this portfolio has managed to beat the 60/40 and the S&P over an extended period of time from the link here. My gripes are similar, TIPs are not easy to invest in, bonds are a high allocation and the indexes he recommends are not easily trackable here.

My main concern about bonds is really what Investment Moats brought up, which is that bonds have been going through a pretty strong bull markets returning large annual returns for a very long time, and it all seems to have bottomed out from here. So where do we go now from here? Which is essentially the main question of the iShares blog post.

So, to end this off, let's have a link from the ever logical thinker from Fabian Capital. Let's have some of my key personal takeways from his article.

"In my opinion an individual investor's portfolio should represent almost the same individuality as their own fingerprint."

I love it how he says that basically, it's up to you, haha!

"It should epitomize your ultimate end goal of a secure retirement, but also embody just the right mix of adaptation to the current market environment."

But, he does emphasize on long term total returns, and promotes portfolio flexibility

"However, if you find yourself wanting to shift your allocations based on market changes, your job has just begun. An active investor should always enact a set of rules to determine which model they should exploit at any given time. Having three to five asset allocation models you can target as baselines to expand or collapse your various sleeves will enable changes to be made more dynamically. Your rules might be based on technical or fundamental analysis, momentum, sentiment, or even astrology. It truly is up to which investment discipline you align yourself with, but the key is to establish a plan, then adhere to it."

And this is where he really shines with a great illustration. If you want to alter your asset allocations, the rules to when and by how much to change the allocations should already be thought up beforehand with some sort of measurable metric, so that we don't change our allocations on whims and fancy, or market hearsay.

This article has excerpts from another article that he has written which has encouraged me to come up with a proper asset allocation model, which can shift its allocations around and adapt according to market factors.

With all that in mind, I think I will be looking towards crafting my own asset allocation model after both Rob Arnott's portfolio and James Picerno's GMI. Definitely a good starting point, along with the traditional 60/40 reinforcement by the iShares research. I will come up with various scenarios, my asset allocations, as well as my rationale for them!

Hopefully, I can also give ETF / fund recommendations to build this complete portfolio too!

No comments:

Post a Comment

Observe the house rules.