SIF Portfolio: Up 18% and still beating the market in year two

The SIF portfolio was launched on 19 April 2016, so last week it turned two years old. The SIF Fantasy Fund stats show me that the portfolio has delivered another market-beating performance last year:

• SIF portfolio 2017/18: +18pc

• FTSE All-Share 2017/18: +3.8pc

• Outperformance: +14.2pc

The portfolio delivered a 19% gain in 2016/17, so last year’s performance is very consistent. (If you’d like to know more about how I select stocks for the portfolio, start here)

SIF’s return over two years is now about 40%. The FTSE All-Share Index has risen by about 16% over the same period. So the portfolio has outperformed its benchmark by 24% in two years.

To flesh out the performance data provided by the Fantasy Fund, I also keep a spreadsheet that enables me to track the performance of each stock in a little more detail. This includes both dividends and costs.

According to my calculations, the average total return for each stock sold since the portfolio’s inception is now 19.8%. This gives an average annualised total return of 31.1% for each position sold to date.

I’m reasonably happy with these figures. Although there are investors on this website who have done a lot better than this over the last two years, the concept behind this portfolio is that it’s largely rules-based and doesn’t require much management.

Avoiding losses

I invest some of my own money in all of the SIF stocks. One of my main goals is to avoid big losses. So I was pleased by the way the portfolio held its value during the first quarter of 2018, when the wider market was falling:

Of course, two years is a short time. And during this period we’ve enjoyed rising markets with low levels of volatility. Like many portfolios (and some professional fund managers), SIF hasn’t yet been tested in a bear market or a recession.

I wrote earlier this year about how I hope my screening rules would perform in more testing scenarios. It’s now time for my annual review of these rules. Can I make any changes to increase performance or provider greater protection against losses?

Did last year’s changes work?

Last year I found that on average, stocks rose after they were sold from the portfolio. So in April 2017, I decided to extend the portfolio’s minimum holding period from six months to nine months. The statistical support for this decision was flimsy, but there were also two more tangible reasons to make the change:

1. The longer holding period meant that every company would publish at least one set of accounts during its time in the portfolio. This refreshes the data in my screen and increases the opportunity to benefit from earnings upgrades.

2. Trading costs would be lower.

To try and gauge if this longer holding period is working, I’ve crunched some numbers. The table below shows the total return on holdings sold after nine months and the percentage of that gain achieved after the six-month point:

So over the last year, holding stocks for an extra three months may have added about 35% to my returns. I accept that this is pretty flimsy statistical evidence. But as far as these figures show anything, they seem to suggest that the longer holding period is working well.

I don’t plan to change my selling rules this year. I’ll continue to hold stocks for a minimum of nine months. They’ll then be reviewed monthly and sold if they no longer pass my screening rules.

What should I change?

There seems to be a consensus view that the market could be more volatile over the coming year. In the US and UK, interest rates are rising. Both countries’ central banks are starting to think about unwinding quantitative easing. If these trends continue I’d expect some knock-on effect on stock market valuations.

Now seems an even better time than usual to focus on balance sheet strength and quality. I’m reasonably happy that my existing rules pick up many issues in this area. The Piotroski F-Score in particular seems to be a good filter for companies with weak finances.

But I think there’s some room for improvement, so I’ve decided to tighten my requirements slightly this year.

Balance sheet strength

Debt: My screening rules currently allow net debt of up to five times net profit. I’m going to reduce this limit to four times net profit. With interest rates rising, I believe highly-leveraged companies should be looking to reduce gearing.

Price/Book value (P/B) > 0: I’m going to add a new rule that excludes companies with a negative book value. This is quite unusual, but should rule out companies with outsized pension deficits, such as Hogg Robinson. This firm has cropped up in my screen several times.

I was tempted to introduce a rule requiring companies to have a tangible book value of greater than zero. But in the end I decided that this might be too severe. I think some asset-light companies can justify operating with negative net tangible assets, if they are sufficiently profitable and cash generative. In my view, examples include portfolio stocks Alliance Pharma and Jardine Lloyd Thompson.

Defensive vs. cyclical

One of my original goals in 2016 was to achieve a roughly equal split between defensive and cyclical stocks. This has proved impossible. Most defensive stocks have consistently been too expensive or too slow growing.

As a result, the portfolio is heavily biased towards cyclical stocks. I believe these have been added at attractive valuations and so far this strategy has worked well. This bias could still hurt the portfolio, but I’ve come to the conclusion that a 50/50 split isn’t practical within my screening rules.

So although I’ll continue to add defensive stocks when I get the chance, I’m officially abandoning this goal.

Thank you

After two years the SIF portfolio has become part of my life and a significant part of my personal investments.

I’ve learned a lot from your comments over this time, so thank you for taking the time to follow this experiment and share your views.

I’ll be back next week with my usual month-end review and an update on the progress of the International SIF.

Disclosure: Roland owns shares of Rio Tinto, Alliance Pharma and Jardine Lloyd Thompson.