The Investment Column: Underperforming high yielders discredit the O'Higgins theory

High-yield stocks have underperformed the market by a wide margin this year, despite conditions that might have suggested the reverse - principally, low interest rates which, all other things being equal, would make shares paying a high dividend relatively attractive.

BZW, which recently studied high yielders, calculated an unusually and unexpectedly high average yield premium to the market.

At this stage in the economic cycle, the broker claimed, relatively strong growth ought to make dividend cuts less likely, negating the need for high yields to compensate.

Yield premiums of the current order are also more usually achieved in periods of high inflation and interest rates.

There are two main reasons this anomaly might have come about. Worries about possible attacks on the level of ACT (which institutional funds can claim back) might be expected to hit shares for which income was perceived to be a greater part of overall return.

Secondly, the high level of recent special dividends and share buy-backs has meant that many funds have been less dependent on high-yield stocks for income. Faced with big cash inflows institutions have shied away from the greater risks implicit in high yielders.

All that has been bad news for two high-income portfolios we created at the beginning of the year using the so-called O'Higgins theory, which gained prominence a couple of years ago after years of steady outperformance and which has equally steadily underperformed ever since.

The O'Higgins share selection technique, adapted for the British market from its US origins, is briefly this:

Take the 10 highest yielding stocks in the FT-SE 100; of these take the five with the lowest share price or, if you prefer, the five with the lowest market capitalisation, hold them for a year and repeat the exercise.

According to the theory, these stocks should outperform because they are unfairly out of favour (hence the high yield) and due for a bounce, relatively safe (in the FT-SE 100 index of leading shares), but small enough within that universe (low market value or, more crudely, low share price) to be able to grow meaningfully.

As the two tables below show, the theory has been better in the abstract than in reality. We created two portfolios, one on the basis of low share price and one using low market value, neither of which has outperformed the All Share index so far this year and one of which has underperformed quite badly.

Looking at the shares chosen by the system it is easy to see why the technique might work. P&O, the best performer, did indeed bounce back from an unnecessarily gloomy market assessment. The announcement of a heavy disposal programme and the expectation that the sales would ensure a maintained dividend payout gave the shares a useful boost.

But elsewhere there is plenty of evidence of the system's potential pitfalls, that shares sometimes have high yields for a good reason.

Hanson's appearance at the head of the high yielders at the end of last year has been justified by the terms of its four-way demerger this year which will almost certainly see a large cut in shareholders' income.

National Grid has been hit by the electricity industry regulator's recent transmission price review and British Gas continues to be plagued by its onerous take or pay contracts.

BZW believes high yielders may be set for a return to favour, especially now we have seen the best of the period of buy-backs and special dividends, but the O'Higgins formula has a lot of catching up to do to regain its former credibility.

High yield dogs still waiting for their day:

Share price performance since 31 Dec 1995

O'Higgins portfolio using low market value

Thames Water -1.5

P&O +25.4

General Accident -1.0

British Steel +17.8

United Utilities* -5.0

---

Average +7.2

FT All-Share +7.7

* Was North West Water

O'Higgins portfolio using low share price

British Steel +17.8

Hanson -16.2

National Grid -15.3

British Gas -21.5

BT +6.2

---

Average -5.8

FT All-Share +7.7