Comparative Tables: The Hidden Charges
1. Introduction
The FSA first made a proposal for Comparative Tables in its October 1999 Consultation Paper Comparative Information for Financial Services. Ten indicators were suggested (in paragraph 6.4), but the Tables contain only "Charges and deductions". These are projections. The Tables are intended to help people choose new products. They do not inform someone who has already bought a product how much it is currently costing per annum.
The Tables list what are known in the industry as "packaged products". The charges and deductions are calculated using the Reduction in Yield (RIY) indicator, although its value is not given in the Tables. In particular, the Tables do not list unit trusts, like the Financial Times websites. These specify management charges and the Total Expense Ratio (TER).
The value of the RIY indicator has to be taken on trust, whereas the TER indicator is calculated from accounts. In the case of unit trusts these accounts are always readily available in booklets.
There are various levels of disclosure of information about funds, according to whether information is contain in:
| 1) | comparative tables, | |
| 2) | fact sheets and key features documents, | |
| 3) | financial statements or accounts available in booklets for each particular fund, | |
| 4) | available only in books combining the information for several funds, | |
| 5) | not available in any of the above. |
Many unit trusts have converted to open-ended investment companies (OEICS), or to be more exact, sub-funds of OEICS. On asking for the accounts investors are sent a report containing the "short form" accounts of all the sub-funds in the OEIC. These reports state that "long form accounts are available". But this apparently means in a lengthy book containing the long form accounts for all the sub-funds.
It is not possible to calculate for example, portfolio turnover from the short form accounts because they do not include total purchases and sales during the year. So that portfolio turnover has gone from position 3) for unit trusts to position 4) for OEICS.
Each sub-fund in an OEIC does not have a legal existence, separate from the other sub-funds. It is the OEIC which has accounts, rather than each particular sub-fund. This is a problem with stakeholder pensions. Those sold by insurance companies are invested in insurance funds, which do not have a legal existence separate from the other funds of the insurance company. Proper accounts are not available on request for each fund. Not only are dealing charges not included in the cap on charges, but it is difficult to find out what they are.
2. "a misleading impression of the price of investing"As in the case of the stakeholder cap on charges the Tables do not include dealing costs. The February 2000 Occasional Paper of the FSA The Price of Retail Investing in the UK by Kevin R. James states:
The Consultation Paper suggested that a new "price indicator" could be developed which includes dealing costs.
"There are also merits in including 'dealing costs' in the price indicator. This would mean that any brokerage paid on the purchase of stocks, shares or securities,and any 'spread' between buying and selling would have to be included in the price." (paragraph 6.15)
This proposed new indicator would:
"include commission payments, management charges, custodians' fees, trustees' fees, audit fees and, possibly, dealing costs. Management charges can also include a range of specific charges such as policy fees and fees for particular activities (like switches between funds)." (paragraph 6.11)
The June 2000 Response Paper to the Consultation Paper states that the proposal to include dealing costs in the Comparative Tables did not receive much support. Almost all the listed respondents were from the financial services industry.
This is a problem in this area in general. For example the Treasury's November 2001 consultation paper Delivering Saving and Assets, refers three times to "most respondents". But most respondents were from the financial services industry. Perhaps these respondents thought one way and everyone else thought otherwise. If the government or FSA then acts in accordance with the wishes of most respondents, this sounds very democratic, but they are in fact following the wishes of the financial services industry. Such responses should be divided according to whether or not they are from the financial services industry.
The dealing costs are in the future, and therefore can only be estimated. But this applies also to the charges and deductions, which are included in the total charges and deductions in the Tables. If dealing costs are not included in the calculation of the totals, then information should be provided in some other way especially by specifying current portfolio turnover. The Consultation Document states:
". . evidence from the US suggests that dealing costs are directly related to levels of portfolio turnover, and that a given fund's level of portfolio turnover tends to remain fairly consistent over time." (paragraph 6.15)
The topic of dealing charges paid by occupational pension schemes has become highly topical, especially as the result of the Myners Review. The development of better disclosure of dealing charges for occupational pension schemes, should be extended to the retail sector.
There have been several recent articles in the press; for example City banks accused of charges rip-off in the Guardian, Top fund managers stick to soft commissions and Managers fail to monitor costs in the Financial Times. The Fund Managers' Association has written a report on the recommendations in the Myners Report
In its response to the Myners Report the Treasury said:
Soft commissions should also be prohibited in the retail sector, such as for the funds of unit trusts.
The government is hoping that the industry will change without the need for new legislation. It promises a review of the dealing charges of occupational pension schemes in two years' time:
It proposes ten questions which trustees should ask their fund managers such as:
"What action have you taken to minimise transaction costs while still dealing effectively?
Which brokers did you deal with and how did you select them?
. . please list other services that you buy or benefits that you receive from the broker concerned . . Please explain how you evaluate the benefit these generate for us relative to the cost.
Explain your rules on entertainment of your staff by brokers, and those with whom you transact on our behalf where we bear the cost."
The same questions should be asked by the trustees of unit trusts and stakeholder schemes of their fund managers. But most stakeholder schemes and other collective investments do not have trustees to ask such questions.
3. Portfolio turnover
In his Occasional Paper Kevin James estimates the "cost of a round trip trade" to be 1.8% ( = 180 bp - "basis points") of the capital traded (Table 2, page 23):
The Cost of a Round-Trip Trade
| Commission | 30 bp |
| Bid/Offer Spread | 75 bp |
| Price Impact | 25 bp |
| Stamp Duty | 50 bp |
| Total | 180 bp |
|---|
These dealing costs are not visible to investors because they are absorbed in the buying and selling price of shares to fund managers. It can be seen that they are not included in the Total Expense Ratio.
If someone is buying a substantial quantity of shares, this tends to increase the market price. This is the "price impact". Some trading systems such as SETS and E-Crossnet permit investors to trade directly with each other and confidentially. This eliminates the bid-offer spread and most of the price impact. There was an article about E-Crossnet recently in the Financial Times by the chief executive Nigel Foster.
The above cost of a round trip trade implies that, as a percentage of capital, dealing costs are at least approximately 1.8 multiplied by portfolio turnover. Where portfolio turnover is defined by T = 100 x (lesser of total purchases and total sales for a year)/(average of capital at start of year and at end of year) %. The Tables should include information about portfolio turnover for all products.
T varies widely between different unit trusts - even those of the same company. One might have expected that a unit trust company would decide that say, T = 50% is about right, and then all its unit trusts would have a turnover of about this value. Natalie McGrane, Project Manager at Fitzrovia, is writing a report analysing the portfolio turnover of UK-based unit trusts and oeics (Natalie.McGrane@Fitzrovia.com), which is expected to be published in November.
The turnover T always seems to be substantially higher than the turnover of the units of the trust. In one particular example, the total net assets are £157 million. The "cash received on creation of units" is £3.7 million and "cash paid on liquidation of units" is £12.1 million, but total portfolio purchases and sales for the year are £150 and £160 million.
If the purchase and cancellation of units is steady, the amounts received on the creation of units can be used to pay for the cancellations. The turnover of the underlying portfolio T is not effected. For these reasons the turnover of units seems to have hardly any effect on the value of T.
The average length of time shares are held for is the reciprocal of the turnover, which is only one year for T = 100%. The total stockbroker's commission for the year is not given in the accounts of unit trusts. High turnovers give the superficial impression that fund managers are doing their job. But. it is not worthwhile doing much research into a company if you are expecting to hold its shares only for one or two years. An article in U.S. News (9th June 1997) describes this as "playing the market" rather than investing. It shows that turnover should not be greater than 20%, and funds with high turnover perform worse. A more recent article in the Financial Times asks just why are portfolios turned over so often?
Dealing costs and portfolio turnover are not mentioned in the Comparative Tables. Savers are being left in the dark.
4. What is included in the calculation of the "charges and deductions" column?
Is anything else not included in the calculation of the charges and deductions of the Tables - in addition to dealing costs? The June 2000 Response Paper to Consultation Paper, states:
"The FSA has concluded that there is insufficient merit in any of the alternatives to justify the use of a different measure in the Comparative Tables from that used in the disclosure regime, and will therefore measure long-term price of products using RIY," (paragraph 6.15)
The charges and deductions in the Comparative Tables are calculated using the Reduction in Yield (RIY) indicator - not specified in the Tables. How is the RIY calculated? The Consultation Paper states:
"At present the PIA uses the measure 'Reduction in Yield' (RIY), but there are variations in the types of costs that are included in the calculation of RIY. For example, custodians' fees and trustees' fees are currently in unit trust RIY calculations, but not those for life insurance products." (paragraph 6.13)
The RIY indicator is defined in the FSA's Conduct of Business rules. These apparently do not say that in the case of life insurance products custodians' fees and trustees' fees are not included in the calculation of the RIY. It is therefore unclear whether or not they are included in the calculation of the charges and deductions column of the Comparative Tables.
In the case of unit trusts they are included in the calculation of the RIY according to rule 6.6.23 which refers to the following rule 6.6.67. But this does not apply to life insurance products:
"The following are those expenses and costs of investment that firms should take into account when making their calculations. The list is not comprehensive. These are in addition to explicit charges: | |
| Examples of expenses are: | |
| (a) registration fees; | |
| (b) safe custody fees | |
| (c) trustees' fees; | |
| (e) audit fees; | |
| (f) regulators fees and subscriptions; | |
(g) costs of investment management but excluding dealing costs of the underlying portfolio, and costs associated with routine management and servicing of existing property investments; | |
| (h) bid/offer spread in the pricing of units | |
The spread in (h) should be on a basis that fairly represents the expected levy of such spread in a firm's experience of normal trading conditions. | |
| The expenses should include allowance for any value added tax which is not recoverable." (COB 6.6.67) |
It should be clear how the charges and deductions are calculated with reference to accounts - like the Total Expense Ratio. A problem is that insurance funds do not have accounts like unit and investment trusts.
A Disclosure Regime should start with properly drawn up accounts, containing information such as total stockbrokers commission, missing from the accounts of unit trusts. Tables are then constructed from information taken from these accounts.
Clicking on the "charges and deductions" in the Tables the explanation states: "This figure just shows the effect of explicit charges and deductions." In the case of life policies the "explicit" applies only to the "charge", not to the "deductions". Insurance funds do not have accounts like unit trusts, where it is possible to see deductions.
The FSA states that the method of calculating the charges and deductions is explained in Chapter 6 of its Conduct of Business Rules, and "Certain parts of the calculation are individual to the product provider."
However the Conduct of Business Sourcebook (Chapters 1-6, February 2001) makes no mention of "charges and deductions". It refers consistently to "charges and expenses", such as:
"The figures in the column 'Effect of deductions to date' must reflect the charges and expenses accumulated at the prescribed rate." (6.5.29 (3))
The prescribed rate for the Tables is 7% per annum. The "charges and deductions" in the Tables is apparently called "effect of deductions" in the Rules.
5. "It's all in"
The Treasury states "Our CAT standards will get rid of the small print and hidden charges that worry people so much.". Looking in the small print of the standards, we find that dealing costs such as stamp duty on dealing in the fund assets are outside the CAT limit on charges:
Thus the CAT standards do not get rid of all hidden charges. They do not get rid of stockbrokers' commissions. As mentioned in Section 2 above, there is considerable concern about the extent of dealing costs for occupational pension schemes. Portfolio turnover for unit trusts and hence for ISAs is substantially higher than for occupational pension schemes. Therefore stockbrokers' commissions will be correspondingly higher. Yet the Treasury is saying that we need not worry about them because they have been got rid of.
An article by a reporter Simon Targett in the Financial Times Fund managers to reveal pension sheme costs (11th March, 2002, page 25) states:
"Under a draft code, to be unveiled this week by the Investment Managers Association, fund managers would provide pension scheme trustees with a six-monthly report detailing the costs of investing pensioners' money on the stock market. For the first time, trustees would know exactly how the £12bn allocated for trading shares, including dealing costs, brokers commissions and government stamp duty, was spent."
Retail investors do not know the cost of trading shares on their behalf, or how this is divided between dealing costs, brokers commission and stamp duty. Occupational pension schemes are leading the way.
Mr Targett reports that "Ministers have set a deadline of March 2003 for the industry to sign up to a code of best practice based on the recommendations of Mr Myners." (15th March, 2002, page 3)
Dealing charges are not included in the stakeholder cap on charges. Whereas for example, the February 2002 Consultative Document of the Inland Revenue Modernising Annuities states that for stakeholder pensions "the annual management charge must be less than 1% of the fund value, with no other charges" (paragraph 27).
The next paragraph of the CAT standards states that stamp duty for dealing in units is included in the charge limit:
Stamp duty consists mainly of "Stamp Duty Reserve Tax" (SDRT), which applies "where there is no written instrument of transfer".
This is explained in a letter about a particular unit trust This states that whereas the liability to stamp duty on the purchase of units was absorbed in the bid/offer spread, it is now paid for out of the fund assets. This seems very unfair to existing unit holders, who have to pay the stamp duty of investors who are buying and selling. Investors are having to pay dealing costs of other investors.
There are thus two kinds of stamp duty, for dealing in the fund assets and for dealing in units of the fund. Neither is included in the calculation of the charges and deductions of the Comparative Tables.
We are constantly being told "it's all in", only to find at a later date, or looking at the small print, that this is far from the case. It happened with the RIY indicator. It has happened with the ISA CAT standards and the stakeholder cap on charges. History is repeating itself yet again. Many people will think that everything is all in the "Charges and deductions" of the Tables. They are not being given the full picture.
6. "We were not given the full picture."Not being given the full picture is given by the Treasury and FSA as reasons for regulatory failure in the case of Equitable Life. In hearings before the Treasury Committee (Regulation of Equitable Life Minutes of Evidence, 30th October 2001), the Minister, Ruthy Kelly, was asked about "the section 68 order on 11 September that rejigged the Equitable balance sheet to the tune of £1.1 billion" (paragraph 152) She replied:
"I do not think this was ever brought to the attention of ministers. . . I do not think we were given the full picture of events." (paragraphs 152,153)
In answer to the question:
"To what extent did the FSA accept without question what the Equitable told it?" (paragraph 25)
Sir Howard Davies, chairman of the FSA, replied:
He reported that successive regulators were not given the full picture:
"It is extraordinary that a company could write so many guaranteed annuity policies and the regulatory returns should contain no reference to the liabilities thereby created." (paragraph 24)
In its April 2001 report Standards for Retail Financial Products the Treasury states:
"In the modern world people will increasingly have to look after their own financial interests for themselves."
This implies an increased need to provide consumers with adequate information about the products of the financial services industry. They are not being given the full picture.
7. Approximately how much do you want to put into the pension?
This question on the Comparative Tables, means contributions net of income tax. The "What is a stakeholder pension" section of the Comparative Tables says "Stakeholder pensions have special tax advantages.":
"The payments you make into a pension receive tax relief, which boosts the amount of your payment. For example, if you are an employee who pays basic rate tax and you make a payment of £100 to your pension, you would currently receive an additional £28.20 tax relief from the government into your fund. The fund you build up also has beneficial tax treatment."
However this is not giving the full picture. It neglects to mention that in addition to tax advantages, there are also tax disadvantages. This often happens with the "products" of the financial services industry - tax advantages are stressed but tax disadvantages only emerge, if at all, at a later date. With a stakeholder pension it is necessary to buy a pension annuity by age 75, as opposed to a purchased life annuity which has more favourable tax treatment. One quarter of capital can be taken as a lump sum, and the rest used to purchase a pension annuity by age 75. The income from the annuity is subject to income tax.
The purchaser of a stakeholder pension (who is a standard rate tax payer) only benefits from the tax concession 28.1/4 = 7%. The same applies for an income drawdown plan. However charges and deductions are increased by the full 28.10% by the tax concession. The pension provider is benefiting proportionately more from the tax concession, which is in some sense an inefficient way of encouraging saving.
8. ConclusionThe "Charges and deductions" columns in the Tables will leave many or most viewers with the misapprehension that they include all charges and deductions. There has been a long history of saying "charges are ...", without mentioning that this does not include all charges; or that there are various further fees, levies etc; which are mentioned, if at all, only in the small print.
Purchasers of the units of unit trusts often do not realise that the management charge does not include all charges. Publicity is given to the management charge, leading investors to believe that "It's all in". The total expense ratio (TER) is a more accurate measure of total charges. It is calculated from the accounts of unit and investment trusts, where the expenses generally appear as "fees".
Insurance funds do not have such accounts. The RIY measure extends the TER concept to life policies. But as we have seen above, it does not include all expenses. Personal and stakeholder pensions sold by insurance companies are legally insurance policies, which do not have accounts like unit trusts, so it is not possible to calculate a TER.
Many viewers of the Tables will miss the adjective "explicit" in the text on Charges and deductions, and will think that the Charges and deductions column includes all charges and deductions. This column should be recalculated to include all charges, expenses, costs etc.
In his 1999 budget speech the Chancellor, Gordon Brown, reported that he had asked the FSA to prepare league tables of the costs and charges of the products of the financial services industry. The Tables are apparently the result of this request. But he did not say in his speech that dealing costs should not be included in the calculation of these Tables.
All possible information on the products of the financial services industry, charges, portfolio turnover, investment performance and so on, needs to be put into a computer system. Then programs run to find out what is going on. For example, when we start to do this for the portfolio turnover of unit trusts, it is immediately apparent that there are substantial differences between turnover according to where the trust is invested, UK, Europe, America, Japan etc. This topic will be discussed in the forthcoming report of Fitzrovia mentioned in Section 3 above.
Such a system would be able to show the association between the amount of portfolio turnover in different years, mentioned above. It would be able to quantify the extent to which past investment performance is a guide to future investment performance. There would be no need to refer to obscure studies in the United States, with the temptation to select only those which support a particular point of view.
Any comments on the above will be appreciated by:
Stephen Wynn, E-mail: centre@boltblue.com
Started 28th November 2001, last update 29th March 2002.