Select Committee on Treasury Minutes of Evidence

Memorandum submitted by Collins Stewart Limited

  To help the Committee this note:

    —  Gives the background to Splits and Zeros.

    —  Examines the impact of changing market conditions on Splits.

    —  Discusses cross holdings and fees.

(a glossary of technical terms is at the end of this memorandum)


  Splits are Closed End listed investment trusts. Investment trusts typically trade at a discount to NAV and Splits are designed to reduce or eradicate this discount. Splits are not new, they have been around since the 19th century. Over time, the Splits have evolved to include new classes of shares such as Zeros shares which carry no dividend or income but entitle the holder to share in the assets of the fund on maturity. In addition to the Zeros, many Splits took advantage of low interest rates by borrowing to provide additional funds for investment to enhance returns (see Annex 1).


  With the changing market conditions in and after 2000, Splits faced the unforeseen circumstances of falling interest rates and falling markets (a most unusual combination), declining dividends, falling ratings and the cost of bank repayments. As liquidity vanished and confidence was lost the prices of shares in Splits moved from a premium to a discount (see Annex 2).


  The changing market conditions had a greater impact on Fund of Funds Splits as the cross holdings, which originated from a desire to spread the assets in Splits, tended to intensify the effect of the falls in equity markets discussed above (see Annex 3).


  The fees and charges involved in launching and annual management are not excessive in comparison with collective investment schemes and new issues (see Annex 4).

25 October 2002

Annex—Background to Splits


  A split capital investment trust is a version of an investment trust company issuing various classes of shares.

  Splits are Closed Ended funds, which (subject to re-structuring) have a fixed pool of assets at the outset represented by a set number of shares in the trust. Shares in the trust should rise or fall in line with their dividends and the NAV underlying each share class. As shares in Splits cannot be redeemed they are listed, allowing investors to realise their investment by selling their shares.

  Splits may have high levels of gearing (achieved either through the relationship between the different share classes or bank borrowing or a combination of the two).

  The main types of share classes are as follows:

Income Shares, offering the potential for high income through dividends, but little or no capital growth;

Capital Shares, not paying dividends but (normally) participating in the assets remaining after repayment of any bank debt and other prior classes of shares, for example Income Shares and Zeros;

Ordinary Shares, offering the potential for high income (through dividends) and capital growth, but ranking after the repayment of any bank debt and other prior classes of shares (eg Zeros); and

Zeros do not pay dividends, instead offer a fixed level of capital growth on redemption provided the Split has enough net assets.

Splits, like other investment trusts, have an investment remit governing their investment policy including investment in other Splits and investment funds (a Fund of Funds)


  Splits have been around for a long time. For example, Foreign & Colonial launched a Split in 1868 which had a fixed life and four share classes, the Hawaiian Investment Company launched a Split in 1880 with £20,000 of share capital and borrowings of £80,000, giving a gearing level of 5x. These were typical structures.

  In the late 1960s, Splits were used as tax effective vehicles, for example income shares would often be suitable for non tax paying investors and capital shares for tax payers whose return was taxed as capital gains at a lower rate.

  In the late 1980s tax treatment of income and capital had come into line. Several large generalist trusts developed the concept further splitting their ordinary shares into income shares, capital shares, stepped preference shares and zero dividend preference shares to reduce their discount to NAV. Their alternative would have been to convert into an Open Ended collective investment scheme.

  Zeros came about at this time to increase the yield available on the income shares to make up for the lower level of income from the equity market when compared to the 1960s.

  The number of Splits started to grow in the late 1980s/early 1990s with existing investment trusts re-organised and many new trusts launched with structures including Zeros.

  Splits based on Fund of Funds were created to broaden the diversification of investments

    —  Most Splits were invested in UK high yield shares which suffered badly in the recession of the early 1990s, for example the NAV of the capital shares in Scottish National (an investment trust with a split capital structure) fell from 300 pence at launch in September 1987 to nothing, but were finally repaid at l04 pence in 1997. The general cut in dividends throughout the market fed through to Splits and many in turn had to cut dividends. Scottish National cut the dividend on their income shares twice.

    —  In the second half of the 1990s, the number of Fund of Funds Splits grew and the sector performed well. Mindful of the problems of the early 1990s managers sought to diversify their portfolios both in terms of asset exposure and in terms of manager risk. This led to the creation of a number of Splits investing in asset classes outside the UK high yield shares for example technology, European shares, US equities and property.

  Asset classes (other than UK high yield) tended to be characterised by low yields. In order to maintain an attractive yield on the highly geared ordinary shares the balanced portfolio Splits, also known as Barbells, were created. These separated their portfolios into growth stock and high yielding securities, which were either corporate bonds, other Splits or a combination of both.

  By the late 1990s, the cost of bank borrowing became relatively cheap compared with the cost of Zeros, leading to a decline in the proportion of Zeros in split capital structures and an increasing proportion of bank borrowing. Some Splits moved away from Zeros completely.

  The theory was that the income from corporate bonds would cover the cost of bank borrowings and the capital growth on the ordinary shares would meet the redemption costs. This worked well as equity markets continued to rise. For example, the ordinary shares of The Technology & Income Trust Limited, launched in July 1999 at 100p, rose to an NAV of 365 pence and a price of 307 pence respectively on 10 March 2000 (they have since fallen to zero as the technology markets in which the trust invested declined).


  The introduction of bank debt into Splits changed the nature of the risks involved in the structures. In the late 1980s and early 1990s, simple structures with up to 50 per cent ordinary shares and 50 per cent Zeros were not unusual. These Zeros tended to have low cover ratios (that is their final entitlement was often greater than the current gross assets of the trust). Also the assets in front of them were only lx their current entitlement (ie they have £50 million of ordinary shares in front of their £50 million current zero entitlement).

  As equity markets grew progressively over the next 10 years, Zeros were increasingly becoming better covered typically with a positive cover ratio of around 1.2x (for example, the final value of the Zero had to be covered 1.2x by the current assets of the trust increasing the likelihood of the investor being paid in full). This was achieved by a combination of reducing the life of the Zeros and reducing the proportion of Zeros in the structure.

  Zeros in newer Splits have therefore been better covered at launch than 10 years ago. A typical Split in the late 1990s could have a 40/20/40 structure, where the ordinary shares represent 40 per cent, the Zeros 20 per cent and the bank borrowings the remaining 40 per cent. The Zeros would have a higher positive cover ratio than the older Zeros. They were better covered under normal market conditions and under conditions of assumed market falls.

  Looking at the two previous examples with the Zeros compounding at 9 per cent for five years, their final values would have grown by 54 per cent. For the old style Split with 50 per cent in Zeros their original £50 million would grow to £77 million. Therefore the gross assets of the trust could fall by 23 per cent over the life and the Zeros would still receive their final value. It would require the assets of the trust to fall to nothing before the Zeros received no return at all.

  For the newer structures including bank debt, the Zeros would rise from £20 million to £31 million. Therefore the assets of the trust could fall by 29 per cent before the Zeros did not receive their full final value.

  However, the crucial difference in our example between the older Zero and the newer Zero is that if the assets fell by 60 per cent the newer Zeros would receive nothing. It is therefore fair to say that in normal market conditions, or in conditions where markets fell by around 25 per cent-30 per cent over the life of the trust, the newer Zeros were less risky. It is also fair to say that in extremely adverse market conditions, such as we have recently experienced, the newer Zeros were more risky than the older Zeros.

Annex 2—Changing Market Conditions

  The continued decline in the value of the shares of leading British companies which represented the bulk of the investments made by Splits assets in the split sector triggered repayment obligations under banking covenants given by the trusts. The technology markets collapsed (Nasdaq fell—77 per cent from peak to trough) and markets generally declined (FTSE 100 fell—46 per cent and the S&P fell—48 per cent peak to trough). This impacted sharply on asset values, which fed through to the Fund of Funds Splits.

  Newer Splits were designed to withstand market falls of 25 per cent-30 per cent but even they suffered.

  Splits faced the unforeseen circumstances of falling interest rates and falling markets. Historically, in a falling interest rate environment, equity markets rise. However, in recent times as interest rates have fallen, markets have also fallen (a most unusual combination). As interest rates fell, the cost of repaying fixed rate debt increased. Splits were faced with the dilemma of repaying bank borrowings and eroding the assets or waiting for markets to recover. As debt is repaid it typically reduces the dividends and the future prospects of the ordinary shares, but may help to underpin the Zeros. As liquidity vanished and confidence was lost, many Split trust prices moved from a premium to a discount. Splits were therefore impacted by falling asset values, declining dividends, falling ratings and the cost of repaying debt.

Annex 3—Cross holdings

  The level of cross holdings in Splits varies enormously between different trust companies. Some have no exposure whilst Fund of Funds Splits clearly have 100 per cent exposure.

  The Fund of Funds concept is designed to provide additional diversification. In the early 1990s, the Fund of Funds Splits were able to maintain their dividend payments despite the dividend cuts by quoted companies. At the time, confidence in Splits was high and their share price volatility low. The expansion of Splits in the late 1990s enabled the Fund of Funds Splits to broaden their diversification through the growth of the sector as new issues began to move away from the traditional UK high-yield index (the FTSE 350 high yield index) and focus on sector themes and geographical diversification (eg technology, property, and US, Far East and European equities). The entry of other fund management groups into the Splits market also served to reduce the exposure to any one fund management group and provide a further means of diversification.

  We show below three charts which illustrate the impact of this.

  The first shows the 10 year performance of the Datastream highly geared ordinary share total return index, compared with the total return of the FTSE All Share index. This shows that the sector returns have been remarkably similar to the equity market until the second half of 2001.

  The second chart shows the relative performance between those two indices. This chart shows the Datastream index outperforming the main market strongly in 1992-93, with the UK equity market catching up in 1996 and the indices performing in line until 2001. The third chart shows the total return on Dartmoor Investment Trust relative to the total return on the FTSE All Share index. Dartmoor is generally used as an example of a Fund of Funds Split, it was one of the earliest and largest. Again the chart shows dramatically that for much of the period shown the trust outperformed the equity market, again until 2001.

  In other words, for all but the most extreme adverse market conditions, Splits performed well.

  Fund of Funds Splits were impacted by falls in equity markets, the additional gearing they had taken on, but also by a significant de-rating in the split shares in which they were investing as discussed in Annex 2).

Annex 4—Fees and Charges

  It is important to separate fees and charges into launch costs and annual costs.

  Launch costs vary on the size of the trust. Typically larger trusts bear a smaller amount of costs as the fixed costs are spread over a larger trust and, in general launch costs could be between 1.5 per cent-2.5 per cent of the gross amount invested.

  Any commissions paid to IFAs and brokers would be additional to the launch cost. Commissions, payable to authorised intermediaries, are not that frequent but if paid are in the range of 1 per cent-3 per cent. The bulk of the regulated institutional investors would not take the commission but would subscribe net of it for the benefit of their clients. Some management groups also underwrote the launch costs at around 2 per cent, such that if the issue was not as successful as hoped the management company and not the trust would pick up the difference. Conversely, if the issue was very successful the management company would receive the benefit of the difference between the actual costs and the 2 per cent charged to the trust. This underwriting provided greater certainty to investors as to their starting net asset values. These costs are low compared to the launch costs for collective investment schemes, which can and are often substantially more and when compared to the fees charged for new issues in the equity market, which can be 5 per cent or more.

  Annual charges would include fees for the fund managers and also all of the other associated costs attributable to the trust, eg fees for directors, auditors, administrators and custodians. In line with the normal market practice, the management fees are charged on the assets managed, and would vary between 0.75 per cent-1.25 per cent of funds under management. The other costs would normally be less than 0.5 per cent which would be paid by the trust company. This would add up to a total expense ratio of around 1.5 per cent per annum of the gross assets managed.


Barbells  a balanced portfolio fund developed to maintain an attractive yield on the highly geared ordinary shares by separating their portfolios into growth stocks and high yielding securities

Closed End  a fund with a fixed initial capital, unlike unit trusts or OEICs

Fund of Funds  a collective investment scheme whose investment policy is to invest in the shares and units of other collective investment schemes

NAV  Net Asset Value

Zero  Zero Dividend Preference Shares

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2003
Prepared 29 January 2003