Tag: asset allocation

What asset allocation performed best over time?

Reading Time: 7 minutes

“Past performance is not indicative of future results” you say as you read this post. But it can be a tool for analyzing what asset allocation performed the best.

The markets are particularly tough to apply history to, though, because of the wide amount of factors involved. Interest rates, valuations, and broad based sentiment are very hard to control for. Imagine looking at the performance of the market in 1938, but not factoring some how for investor sentiment in a world about to enter WWII.

However, to think we can’t learn anything from studying history is non-sense.

Today, I’ll be taking a look at what performed best over recent cycles in the equity market and walking through some of the take-aways I have. Feel free to share your results as well in the comments! I’ll be using http://www.portfoliovisualizer.com which has some great tools.

My father is currently entering retirement and asked for some advice on how to manage his portfolio. As a retiree, he needs growth, but also needs to weather the downturns a bit better than the normal investor.

As such, I thought I’d take a look and see, “What asset allocation of sectors performed best over the past 2-3 business cycles?

This brings up the concept that I call, “Win by not losing“. Let me explain. A lot of people enter the market thinking that they need to buy high growth stocks and want to ride the tidal wave in order to get rich. And quick.

But as we saw from the tech bubble, that can quickly go against you. While you may make a lot of money early on, if your portfolio declines by 50%, you need to go up 100% to be back to breakeven after that.

I don’t believe in timing the markets, but I do believe a good asset allocation can help performance in a variety of markets.

Alright, let’s head over to portfolio visualizer and I am going to go to their portfolio optimization section. Here, we can backtest how different allocations of sectors would have performed over different time periods. The great thing about this tool is we can also select what we want to optimize the asset allocation by.

Screenshot 2019-02-16 at 4.34.09 PM

We need to select funds that capture each sector of the market. Unfortunately, ETFs haven’t actually been mainstream for that long, so funding sector ETFs that go back to 1985 is impossible. I can go back to 1999 though by selecting these funds:

  • XLV – Healthcare
  • XLF – Financials
  • XLE – Energy
  • XLU – Utilities
  • XLB – Materials
  • XLI – Industrials
  • XLY – Consumer Discretionary
  • XLP – Consumer Staples
  • QQQ – Tech
  • FREEZ – Real estate (no ETFs went back this far, so am using a fund)
  • FSTCX – Telecom (No ETFs with enough history. IYZ goes back to June 2000)
  • SGGDX – Gold (again, not ETFs)

This gets us enough history to go back to January 1999. That’s important because we can capture the end of the tech bubble, the early 2000’s recession, the recovery over the mid-2000s, the boom up until 2007, the great financial crisis, and the bull market since then (while also capturing all the wobbles in between including the european debt crisis fears, the interest rate fears, the commodity collapse, and so on). Would I like more history? Of course, but this will make do.

I will first equal weight all 12 of these funds (e.g. 8.33% each) and seek to find the “maximum return with the minimum volatility”. I will set the maximum volatility at 15% also make it so each fund is a minimum of 2.5% and a maximum of 25% (except gold, I will cap at 15% for practical purposes).

Realistically, we need a parameter like this because we can’t know if energy for example is going to tank in 2015/2016 after being super stable from 2011-2014. Housing during the 2000s too is another example. As such, we have to allow for some diversification.

I am not including global ex-US funds, but you can if you’d like. I may do this as a follow up post.

January 1999 – January 2019

Alright, in your mind, how do you think this asset allocation will perform against the S&P 500? For context, here is the current breakdown of sectors:

Screenshot 2019-02-16 at 4.58.25 PM

How does our equal-weight asset allocation stand a chance without the massive secular growth story of tech?

Well, lets take a look:

Screenshot 2019-02-16 at 5.05.38 PM

Wow… the equal weight portfolio crushed the S&P… Investing $10k in the S&P500 would have turned into $30k, but the equal weight was nearly $45k. But the red line crushed both. It ended with $54k in value.

What in the world did it allocate to?? I have to admit, I would not have guessed this:

  • Real Estate – 25%
  • Consumer Discretionary – 25%
  • Tech – 15%
  • Gold – 15%
  • Everything else – 2.5%

When looking at the annual returns for each asset class it starts to make sense. When tech was taking a beating in 2000 and down 36%, real estate was up 30%, so early on this diversifying factor helped. In 2008, people jumped to buy gold in panic, which also helped. I am a bit surprised the healthcare, staples and utilities were not weighted hire for their defensiveness qualities.

Screenshot 2019-02-16 at 5.17.22 PM

Let’s change the time series.

January 2005 – January 2019

 Why this time frame? Because it feels relevant. The fed is raising rates (as it was then) and it is closer to being late in the cycle than being in the early part of the cycle (as it was then, thanks to my hindsight vision).

Again, I just want you to try to guess what would have outperformed against our equal weight portfolio. Spoiler alert: it crushed us.

Screenshot 2019-02-16 at 5.22.32 PM.png

The equal weight performed on par with the S&P, but the weightings of this portfolio really separated from the pack in around 2014 or so.

Screenshot 2019-02-16 at 5.24.33 PM.png

Ah, now it makes sense. Tech comes into the fold and starts crushing the rest of the market. An overweight in this sector definitely added to the out-performance of late (FANG).

I also think this might be impacting consumer discretionary to a certain degree given Amazon’s weighting in that index (it currently sits at ~21% of the XLY index weighting!). Healthcare performed well in 2008 and then experienced strong growth since then.

Here’s an interesting question… How would our top picks from the last backtesting performed in this back testing? Well, when you take out the performance of real estate early on from the year 2000, it puts it on track with the S&P500 (though it did perform well early on in the cycle).

Screenshot 2019-02-16 at 5.35.31 PM.png

Let’s take the gloves off. Allow me now to show the results of an asset allocation that may surprise you, but will hopefully tie things together. I want to allocate my portfolio in a way that I “win by not losing”. Here’s my allocation selection:

  • 25.0% Consumer Staples
  • 22.5% Healthcare
  • 15.0% Utilities
  • 10.0% Tech
  • 10.0% Gold
  • 2.5% Everything Else

This is called “Portfolio 1” in the chart below and “Portfolio 2” is an equal weight of the sectors. I’ve selected re-balancing annually to keep the weights in check.

Over a long period of time, this crushes the S&P, which we’ll look at. It doesn’t seem to make much difference vs. an equal weight tough.

Screenshot 2019-02-18 at 12.48.23 PM.png

One stark difference though is that portfolio 1’s worst year was much better than the other two. So portfolio 1’s asset allocation protects from the downside. 

Screenshot 2019-02-18 at 12.50.01 PM.png

So let’s look how this performed starting from 2005 instead of 1999.

Here the performance difference is much more stark and you can start to visually see why. Take a look back at the most recent sell off in Q4’2018. Portfolio 1 barely ticked down while the equal weighted and S&P500 suffered greater losses.

Screenshot 2019-02-18 at 12.52.32 PM

Last but not least, lets say again you think we are late in the cycle, but want to stay invested to meet certain fiscal goals. The analogue here is 2005 – late in the cycle and a recession looming a couple years out. Let’s see how stark the performance is up close, from beginning of 2005 to the end of 2012.

Screenshot 2019-02-18 at 1.00.07 PM

This is the most important slide to me. To me, it highlights that someone who played good defense ended up recouping their losses in the crises much faster than the S&P500 (which still did not make it back to its high watermark over this period).

Obviously, some of these sectors such as Utilities and Consumer Staples have benefited from lower interest rates over the past 30 years and its something to consider for asset allocation going forward. However, in a year when rates moved up considerably, these sectors still provided cover in 2018 (outperforming the S&P’s year by 110bps and suffering only a 7% drawdown compared to the 13.5% drawdown from Oct to December 2018 suffered by the S&P.

I’m going to end with a cliche and say sometimes, the tortoise (consumer staples) beats the hare (insert the “it” growth sector here). And maybe asset allocation should be set up that way too.