Thursday 23 April 2015

'Backtesting' the DIY World ETF Income Portfolios

Does back testing help traders?
Last month I suggested that it would be possible to create a DIY geographically-diversified income-oriented portfolio using just ETFs (Exchange Traded Funds). The idea was to create a long-term DIY portfolio that was easy to administer and which also minimised risk, whilst retaining (what I see as) the benefits of a yield-based income-oriented approach to investment.

One reader asked whether it would be possible to review the past performance of such a portfolio - so here goes...

In fact I suggested two portfolios - a lite version one and a slightly-more-complicated standard one - based on the same principles:
  • worldwide diversification, as far as is possible using ETFs quoted on the London Stock Exchange
  • a rough 50/50 split between ETFs holding dividend shares and fixed-income debt (bonds)

The lite portfolio consists of only two 'worldwide' ETFs (a dividend share one and a fixed-income one), which will never be sold (although they will be added to).

The more complicated standard portfolio, consisting of a handful of ETFs has two additional rules
  • purchase rule: highest yield consistent with geographical diversification
  • sale rule: when yield falls below 3%

These are the kind of DIY income portfolios that you might set up as a family member's Self-Invested Personal Pension - particularly if that person wasn't very interested in the investment process.

The 'Lite' Version

The simplest portfolio of this type  I could come up with consists of just two ETFs:

Big advantage: simplicity. Big disadvantage: low-ish yields. 

Performance (calendar year or Year-To-Date total returns, according to
  • VHYL: 2014:  7.6%; 2015 YTD 5.8%; Current yield: 3.2%
  • IGHY: 2014: -0.4%; 2015 YTD 4.0%; Current yield 4.9%

Assuming equal proportions, the average returns were:
  • over 2014 around 3.6%, which is definitely nothing to write home about
  • for 2015 YTD around 4%, which is actually not too bad for the first third of the year.

But not spectacular, although for this version we are looking more for long-term security, so lower returns might be acceptable. Occasional admin (e.g. annually) would be needed to top-up each ETF with the income received.

The 'Standard' Version

(See the original post for more details)

The 'Standard' version of the DIY World ETF Income portfolio would consist of (currently) six ETFs that satisfy the yield criterion:

Performance (calendar year or Year-To-Date total returns, according to
  • IUKD: 2014:  7.0%; 2015 YTD 6.8%; Current yield: 4.3%
  • SHYU: 2014:  14.7%; 2015 YTD 7.3%; Current yield 5.2% (Bloomberg data)
  • SHYG: 2014:  -7.9%; 2015 YTD -4.6%; Current yield: 4.7%
  • IAPD: 2014:  -1.8%; 2015 YTD 4.5%; Current yield: 5.8% 
  • SEDY: 2014:  -7.5%; 2015 YTD 3.5%; Current yield: 4.9%
  • SEML: 2014:  -5.0%; 2015 YTD 1.5%; Current yield: 6.0%

    Here's the disappointment: the returns on this collection of ETFs over the last year and a half are certainly not very exciting, with the exception of SHYU. I'm not even going to try to calculate the theoretical return. So does this mean that the DIY Worldwide ETF Income strategy is doomed to fail?

    Having now gone through this exercise, I can see the problem in trying to backtest this strategy. The key point about investing in high yield is that the yield is high (usually) because the price (and therefore total returns) have fallen, because they are out-of-favour with the market. We buy high-yield securities in the hope that the market will reappraise them and consequently the price will recover - giving a total return of the high yield plus the price increase.

    In other words, looking at the historical return of something that is currently high yield (because it has experienced a price cut) is obviously not going to reveal a very attractive historical return!

    As the prices (and yields) change, and the income builds up in the portfolio, it will be necessary to buy some new ETF - either topping-up existing holdings or trying something new (and high-yield). It may also be necessary to sell an ETF whose price has risen to the point where its current yield is no longer attractive (i.e. less than around 3%). Intervention is still required, so I have only partially solved the challenge of a low-admin DIY income-oriented portfolio. 

    So, unfortunately, the conclusion is: if you want to see how this ETF selection has done you'll have to wait until next year for a review!

    But a more interesting thought is that maybe I will have to start looking at a completely new approach: using income-oriented Investment Trusts (which I have not invested in before because of their high charges and obscure 'black box' operations) to create a completely automated SIPP (or ISA).

    I am not a financial adviser and the information provided does not constitute financial advice. You should always do your own research on top of what you learn here to ensure that it's right for your specific circumstances.

    1 comment:

    1. Investment Trusts, now that's more like it, you now have my attention !

      Although you have concerns about their charges etc. I think that overall if you do the research and select the right ones along the same lines as you have done with the high income/yield and regional ETF's then I believe that you may get better returns on Inv. Trusts than the YTD has shown for your ETF's thus far.

      There are lots of resources and blogs out there, here are a few.
      for analysis and comparison of Investment Trusts.

      On the Motley Fool boards, there is always active discussion on
      the Investment Trusts & Unit Trusts boards.
      You are probably aware of this site, but for others viewers, John updates his Investment Trust Income Portfolio a few times a year, and previous updates are on the portfolio tab.

      John Baron (who has a monthly column in the Investors Chronicle) has written a great book on Investment Trusts and now updates this portfolio on

      Great post, all the best