Join our community of smart investors

Further Reading: Small caps win despite market’s population problem

UK-listed small companies beat the FTSE All-Share by 6 per cent in 2019
February 20, 2020

Hobby investors love the high-octane world of small-cap shares because compensation for lower liquidity, higher spreads and oversight costs can come in the form of multi-bagger returns. Aggregating small-cap performance on the main UK market, the 2019 review of the Numis Small Companies Index (NSCI), shows this size premium is in rude health. The most widely used benchmark, NSCI XIC – which excludes investment trusts – made 25.2 per cent including dividends in the 2019 calendar year, versus 19.2 per cent for the large-cap oriented FTSE All-Share index.

Cumulatively, based on 33 years of live data plus back-testing to 1955, the NSCI’s performance makes a compelling case for small companies. The review authors, Professor Paul Marsh and Scott Evans of the London Business School, note a hypothetical £1 that tracked NSCI XIC from 1955 was worth £8,688 at the end of 2019, versus £1,181 if that solitary pound had been invested in the All-Share.

 

Are opportunities drying up?

Of course, indices have inherent survivorship bias, and impressive overall figures mask countless tales of failed businesses where investors have been wiped out. Given it would have been impossible to buy an index tracker, the size premium from 1955 can’t be taken to mean loading up with small companies is the best way to guarantee long-run returns. What it does prove, is that there have been big winners and the enticement of small-caps can be more than just sizzle.

Sensible investors who heed the usual risk management caveats and understand sometimes you come off worse living by the sword, might be most concerned by a decline in the number of small-cap opportunities. The LBS academics note a striking feature of the NSCI, and UK and US equity markets generally, is the falling number of listed constituents. Mimicking western demographic trends, a key factor is the birth rate – ie, the number of initial public offerings (IPOs) lagging the ‘death rate’ of companies being taken over or becoming worthless.

The NSCI XIC began 2020 with 346 constituents compared with 1,255 in 1997 and the lack of fresh listings is a major factor in the downward trend. The graph shows the main market has seen a net transfer of 166 companies to Aim, but most importantly, IPOs have failed to keep pace with the death rate. Mr Evans notes “2019 was a very subdued year for IPOs – the third lowest since 1987”.

Since the late 1990s there have been two savage bear markets but, on both sides of the Atlantic, the availability of financing alternatives for companies has been mooted to explain IPOs not recovering strongly. Although they haven’t researched the impact the private equity (PE) industry has had, the academics give credence to the argument it’s a factor.

 

NSCI Stocks now larger and more liquid

There are positive aspects to the changing characteristics of NSCI as its value hasn’t fallen, rather there is greater concentration in fewer companies. This means NSCI shares are larger and more liquid – so easier to trade – although in their review the LBS team assert “a reinvigorated IPO market would be beneficial and healthy for all investors”.

Another interesting feature of the Numis report is acquisition activity was subdued in 2019. Compared with the long-run picture over 65 years, there was little that took place, with just 1.7 per cent of NSCI index value being acquired. Although bid premiums are often high, the report says acquisitions on average add only 1.1 per cent to index performance.

Going forward, it is impossible to predict how the size premium will behave and investors should expect much year-to-year variation. For example, there may exist a perception that in good years for small-caps the smallest companies should do better. The results for 2019 demonstrate it is wrong to make such assumptions. In what was a fine 12 months for NSCI, the minnows included in another index, the NSC 1000 (named for its starting base number, not the number of constituents), did less well and made 15.3 per cent, also under-shooting the All-Share.