New and improved: the IC High Yield System

Yet while the products available to track strategies are evolving and becoming more cost effective, taking a systematic approach in search of optimal returns is nothing new. One needs to look no further than the success of many Investors Chronicle stock screens and portfolios to understand the potential rewards of a rules-based approach. With this in mind, it seems appropriate to revisit a readers' favourite from the past; namely Investors Chronicle's High-Yield System.

First conceived by a stockbroker at the end of the 1930s, the system's guiding principle is that higher-yielding shares perform better over time. The original author put this down to investors' tendency to overestimate the risk of some securities so that good, albeit less certain, investments are often valued on too high a yield. Consequently, as well as higher dividends, owners of these shares benefit from capital gains as the share price rises to reflect the true risk profile relative to the market.

Recognising that some riskier investments thoroughly warrant a high yield, the system incorporated selection rules relating to market value and share price trend to act as a quality safeguard. All shares on the Official List that met the criteria were ranked by dividend yield to give a portfolio of 30 shares. Like many of the smart beta strategies marketed today, the High-Yield System equal-weighted the portfolio so that the total investment was split evenly between all 30 holdings.

Bringing the High-Yield System up to date

The full rules for the 1939 selection process were:

1) the shares must be quoted on the Official List;

2) market value of the equity must exceed £1m;

3) the shares must be receiving dividends; and

4) the price of the shares must be higher than it was a year ago.

These simple rules stand the test of time, although they do require some obvious adjustment in respect of the market value. Selecting a new portfolio for each year between 1994 and 2014 from the FTSE All-Share, I set out to back-test and examine the year-on-year performance of a revamped High-Yield System. I started by screening for shares that a) paid a dividend; and b) had experienced positive share price movement over the previous 12 months.

Deciding upon a market value criteria was less straightforward. It would be incorrect to simply factorise the £1m figure to allow for inflation. A better method would be to adjust the cut-off upward in line with the overall growth of the stock market. In the end I settled on screening according to a percentile of the market values of companies that met my first two criteria; only the top 40 per cent by market capitalisation would make the cut. Such a criteria means we need not worry about the effective of inflation on market capitalisations in future.

A second filter excluded shares that ranked in the top 10 per cent by dividend yield. This may seem counter-intuitive but it is in keeping with the intention of the High-Yield System to select good-quality shares that one might be happy to hold for a longer period of time. After all, unusually high yields can often be a precursor of trouble.

The remaining shares were ranked by dividend yield. Secondary or cross-listings were excluded to remove overseas companies that report dividends before taxation. I did keep investment trusts in the selection as they qualify as UK-listed companies paying a dividend. In a last alteration to the original ground rules, the final selection was shortened to the 15 shares of the remaining list with the highest dividend yield; these were given an equal-weighting in the portfolio.

Market-beating results

The new strategy, which assumed selling the previous holdings and buying the new selection of 15 shares each year, delivered an outstanding performance back-tested to 1994. In terms of price returns, the new High-Yield System made a 249 per cent gain overall, compared with 115 per cent for the FTSE All-Share. As you would expect from a dividend-led strategy, total returns were even more impressive, with the selections delivering 807 per cent versus 309 per cent for the benchmark. That's over two and a half times outperformance or, to put it another way, £1,000 invested in 1994 would now be worth £4,090 had it tracked the FTSE All-Share but would be worth £9,070 had it tracked the High-Yield System.

FTSE All Share vs High Yield System since 1994 price and total returns

Charts FTSE All-Share versus High Yield System since 1994

Source: Thomson Datastream

Examination of the returns by year, demonstrates that the power of the strategy lies in its consistency. The system's outperformance of the All-Share benchmark in 13 years out of 20 is impressive but the key to success over the long run has been resilience in the face of market shocks. For example, the High-Yield System would have delivered positive total returns in 2000-03 while much of the market was getting hammered. Conversely, the system had its worst year in 1999 during the last stages of the dot-com bubble.

An explanation can be found in the quality controls that underpin the selection process excluding many of the more volatile shares of that period. In the 1930s the system's original author based his theory on the premise that high yields were a premium for greater perceived risk. Since then investors have also paid a premium for less risky shares that pay a higher dividend. The modern selection criteria, that excludes companies with a lower market capitalisation and/or very high dividend yields, has led to the inclusion of some of these bellwethers of the stock market and given some of the portfolios selected a defensive bias. The system also performed poorly in the 2007-09 bear market but did less badly than the FTSE All-Share benchmark and, while not rebounding as strongly initially, it has delivered better total returns in recent years too.

Latterly, we have also seen the emergence of smart beta exchange traded products that track indices of high-yielding shares. The iShares UK Dividend UCITS ETF (IUKD) tracks the FTSE UK Dividend+ Index and is one of the exchange traded funds in Investors Chronicle's Top 50 ETFs. This product differs from the new High-Yield System in terms of its selection criteria and weighting - it physically replicates the leading 50 dividend payers of the FTSE 350 and constituents are weighted by one-year forecast dividend yield - but it provides a useful additional benchmark for our strategy. Rebasing to 2007, we see that the High-Yield System would also have done significantly better than the iShares product. Taking 2009 as our starting point, towards the end of the savage post financial crisis bear market, the iShares product has achieved higher returns than the High-Yield System and both have comfortably beaten the FTSE All-Share and the FTSE 350.

FTSE All-Share vs. FTSE 350 vs. High Yield vs. iShares since 2009

Source: Thomson Datastream

A portfolio to run for the future

Of course, the past is not a guide to future performance and looking at returns you might have achieved in hindsight is very different to actually staying invested and maintaining discipline in a strategy. That said, the case for the High-Yield System is compelling and I have created a new portfolio of 15 shares to track going forward. As I intend to run the portfolio, I have added dividend cover to my selection criteria to reduce the risk of a dividend being passed.

The author of the 1930s portfolio intended to conduct quarterly reviews with the yield on each share and the average for the whole group again worked out. Any share in the portfolio would be sold (a) if its yield was less than the average yield of the whole list - provided its price was below that of a year ago; and (b) if its dividend was passed. These rules are designed to make one automatically run profits and cut losses, so will form the basis of my first quarterly review of the portfolio in three months' time. Another advantage is that the turnover of shares, and hence the impact of dealing charges, is reduced. Quite remarkable, when you consider the creator of the original system was a stockbroker.

New July 2014 selection of High-Yield System shares

Source: Thomson Datastream