1. "Most investors would not benefit."

The feedback paper of the FSA Bundled brockerage and soft commission arrangements for retail investment funds (2006) PS 06/5, ^x is not in favour of the disclosure of transaction costs:

"We believe that making the disclosure information publically available would improve the transparency of costs borne by investors. We believe that most investors would not benefit from such information, so it would not be proportionate to require firms to provide it to them automatically. In particular, it is questionable whether the costs inherent in disclosure to investors would be justifiable, given that consumers are unlikely to make effective use of the information disclosed." (paragraph 2.6, page 7) ^x

The terms "disclosure", "publically available", "improve transparency", "provide it to them automatically" seem rather vague. Unit trusts and OEICS have accounts. There will be little or no cost in putting additional information in accounts.

Transaction costs can be estimated in various ways. The First Report of the Pensions Commission says:

"Estimating the implicit costs is difficult but vital, if as seems likely, the implicit costs are for some investors as significant as the explicit costs." (page 214)

If they were disclosed then they would not need to be estimated. They can be estimated in various ways, starting with investment performance.

2. Investment performance

The diagram below "The price of investing", is from Annex 2 of the FSA's paper Comparative Information for Financial Services (1999) (CP 28). ^x It shows unit trusts, from Micropal data, underperforming the market by 3.7 % per annum, of which 1.2 % is due to the annual management charge. So that total costs less the management is 2.5%. Some of this is expenses such as fees to auditors, which are an explicit cost "in the same print" since they appear in the accounts. The implicit cost is then 2.5% less expenses - perhaps 2%.

Implicit costs are shown in Figure 6.10 in the First Report of the Pension Commission. It says:

"One major study published by the FSA in 2000 suggests that they could be as high as 1.3 per cent for actively managed funds, and 0.5-1.3% even for index-trackers. Some of these results look surprisingly high and the Pensions Commission will be doing more work to validate results." (page 216)

The Pensions Commission does not seem to have done further work on this topic since there is no mention of "implicit costs" in its Second Report.

This is referring to the FSA report by Kevin James The price of retail investing in the UK ^x (2000). This uses the same Micropal data mentioned above, over the same time period (but to be included funds need to exist for slightly different durations). It distinguishes between "the simple average method" (3.1.1) which uses arithmetic means and "the econometric method" (3.1.2) which uses geometric means. The calculated costs are higher for the former method. The explicit + implicit costs are respectively 3.16% and 2.71% in Tables 6 and 9 in his paper. We should use arithmetic means but the Pensions Commission has taken the latter method to reach its 1.3% figure for implicit costs. (To obtain 1.3% it subtracted the 1.4% in the first column in Table 5, from 2.71%.) Therefore the 1.3% for implicit costs quoted by the Pensions Commission is arguably too low.

3. As a multiple of portfolio turnover

The First Report of the Pensions Commission (page 218) quotes 1.8% as the cost of a two-way trade from the paper of Kevin James and suggests that 1.48% is a more accurate figure, because of a reduced bid-offer spread cost since it was published. This is shown on Table 1.

Table 1

The Cost of a Round-Trip Trade ( 100 % Turnover )

Commission	0.30 %
Buy/Sell Spread	0.75 %
Price Impact	0.25 %
Stamp Duty	0.50 %
Total	1.80 %

In the Fitzrovia report "average portfolio turnover for UK actively managed equity funds" is 62.3% in 2003. We can then multiply by 1.6 and obtain 1%. This seems a reasonable conservative estimate for the average value of implicit costs are a percentage of assets. It is lower than the 1.3% in Section 2.

Implicit costs are lower on average for trackers, because they have lower portfolio turnover than actively managed funds. The 1.3% implicit cost in Section 2 becomes 0.88% for trackers. The Fitzrovia report Portfolio Turnover for UK Funds (2003) gives (page 23) average portfolio turnover for actively managed equity funds - 72.3%, mixed equities/bonds - 62.0%, index tracking equities - 25.3%.

There are various kinds of trackers. ^x Turnover can be high if there is heavy use of derivatives. The turnover of the Scottish Widows International Equity Tracker is stated by Fitzrovia to be 685.5% in 2001 and 571.7% in 2002.

Average portfolio turnover varies between different companies as shown in the Addendum below. Funds sometimes buy and sell the same shares during the year. For example Artemis UK Special Situations Fund is stated by Fitzrovia to have a turnover of 356.3% in 2001. In the 2005 accounts (page 17) the largest purchases and sales were both Vodaphone. ^x

4. The range of portfolio turnover

Table 2

"Funds by Portfolio Turnover"

Rank- ing	Asset Class	Company	Purch. £ '000	Sales £ '000	Size £	PT %	Yr
1	Mix Equ./B.	First State Funds - Balanced	232,397	975,929	12	1861.7	3
2	Index Equities	Scottish Widows Int. Equ. Tracker	1,091,969	1,060,675	154	685.5	1
3	Equities	ISIS Funds II - UK Prime	176,785	161,802	26	625.4	2
2589	Equities	Scottish Widows Millennium	51	209	11	0.5	2
2590	Index Equities	Gartmore Japan Strategy F.	116	310	27	0.4	3
2591	Index Equities	Gartmore Index USA	128	217	98	0.1	3

Source: first and last three rows of Table 29 in the 2003 Fitzrovia Report Portfolio Turnover of UK Funds

5. Total amounts

5.1 From total assets

Table 2

Assets of institutions (£ billion)

year (end)	unit trusts	investment trusts	insurance companies	pension funds	% pension funds
1993	88	29	434	480	46.6
1998	163	47	776	699	41.5
2000	223	60	933	765	38.6
2002	190	38	854	610	36.1

Source: Financial Statistics , ONS x

We are interested in the retail sector. So that adding together unit trusts, investment trusts and insurance companies £ 190 + 38 + 854 billion = £ 1,082 billion. We can then multiply by 1% and obtain £10 billion per annum as the total implicit costs.

5.2 Total stockbroker commissions

"The disclosure information" in the quotation in from the feedback paper quoted in Section 1 includes stockbrokers' commission. The paper says for example: "We believe that commission spending out of retail funds might not receive proper consideration." (paragraph 1.2)

There are all sorts of estimates in the literature about what are total stockbroker commissions. The recent article Transparency on commissions could transform investment by Philip Augar and Paul Myners says that:

"UK equity dealing commissions paid by UK investors alone exceed £2bn a year, with significant additional costs in respect of international equity investment." ^x

They refer to "costly and frequently value destroying trading". Paul Myners is reported to have said in 2003 "nearly £1 billion of the trading costs was paid as commission to investment brokers." ^x In the same year the Consultation Paper of the FSA Bundled Brokerage and Soft Commission Arrangements (April 2003) ^x said:

"In 2000, for example, UK fund managers paid about £ 2.3 bn from their clients' funds to UK brokers." (1.1)

"In 2000 over 2.3 billion were paid as commission by UK institutional fund managers to UK brokers. It is estimated that fund managers spent between 660 and 880 million of the commission on services additional to dealing." ^x

Also in the same year, the report from OXERA An Assessment of Soft Commission Arrangements and Bundled Brokerage Services in the UK (2003) ^x says that total brokerage commission increased "from £ 1.5 bn in 1992 to £ 5.7 bn in 2000" (paragraph 103) and was increasing by 18% per annum. Of the £ 5.7 bn, £ 4.5 bn was paid by institutional clients as opposed to private clients (paragraph 218).

From Table 1 total transaction costs are 6 times commission costs. Multiply £4.5 billion by 6 and we obtain over £26 billion. This includes a value for pension funds which we are excluding, but it still seems remarkably high. We suspect that the 30% stockbrokers' commission has been understated in Table 1. It is correct according to the OXERA report, Table 2.6.

5.3 Stamp duty

The amount (£ billion) of stamp duty paid on stocks and shares 1998-2004 was: ^x

Year	1998-9	9-2000	2000-1	2001-2	2002-3	2003-4
S Duty	2488.0	3711.3	4476.9	2852.1	2538.1	2559.3

From Table 1 the total implicit costs are found by multiplying stamp duty by at least three. So we obtain roughly £10 billion per annum.

Almost all this stamp duty is "stamp duty reserve tax". But this does not distinguish between two kinds of stamp duty, on the units (in the case of a unit trust) or fund shares (in the case of an OIEC), and on the underlying investment.

6. Soft commissions

Official concern with the transparency of transaction costs was initiated by the report Institutional Investment in the United Kingdom: A Review ^x by Paul Myners. This report is particularly concerned with pension funds, such as paragraphs 58 - 61. The FSA said that in 2000:

"Up to 40% of total commission spend is used to acquire services additional to dealing so it is important that investors are clear on how their money is spent." ^x

The OXERA report (paragraph 22) found that only 6.8% of the commissions in their survey were soft commissions.

7. The need for disclosure

Transaction costs are highly dependent on portfolio turnover because the more dealing the higher these costs.

Investors need to know:

a) portfolio turnover,
b) total stockbrokers' commission,
c) whether or not there are bundled brokerage or soft commission agreements,
d) two kinds of stamp duty: 1) on the units (in the case of a unit trust) and 2) on the underlying investments.

Transaction costs need to be disclosed as far as possible, because they apparently amount to over £10 billion per annum and vary greatly between different funds. Looking for example, at a Statement of Recommended Practice of the FSA for OEICS (CP 36), ^x I cannot find any reference to either transaction costs, portfolio turnover, stockbrokers' commission or stamp duty.

The FSA has published several reports with the title Bundled Brokerage and Soft Commission Arrangements. In the first instance total stockbrokers' commissions need to be disclosed. As it says in its April 2003 report under the heading Public Awareness:

"Increasing the transparency of commission costs will improve the information available to consumers about the costs of investing. This will enable them to assess more accurately the benefits and risks of investing in managed funds and thus to make better-informed and more suitable choices." (A 2.13, page 47) ^x

Addendum ^x

Average Portfolio Turnover by Asset Management Company
Fitzrovia Report 1st Edition

EQUITIES

Ranking Edition 1st 2nd	Straight Mean Portfolio Turnover %	No. of Funds		Company

01 14 02 62 03 09 04 00 05 06 06 18 07 76 08 05 09 31 10 03 11 19 12 00 13 11 14 00 15 00 16 00 17 32 18 23 19 70 20 51 21 25 22 46 23 15 24 10 25 79 26 40 27 00 28 38 29 27 30 17 31 28 32 57 33 49 34 52 35 48 36 30 37 43 38 36 39 47 40 16 41 26 42 80 43 65 44 71 45 08 46 21 47 67 48 00 49 00 50 00 51 29 52 64 53 68 54 00 55 92 56 69 57 56 58 33 59 66 60 93 61 59 62 20 63 82 64 00 65 00 66 85 67 00 68 63 69 83 70 45 71 78 72 90 73 95 74 75 75 84 76 89 77 77 78 39 79 61 80 94 81 00 82 00 83 88 84 46 85 13 86 22 87 91 88 00 89 87 90 00 91 98 92 99

176.2% 162.2% 155.7% 141.1% 140.2% 125.1% 123.7% 109.6% 109.1% 104.4% 102.5% 101.2% 99.5% 96.9% 94.6% 87.0% 82.2% 80.7% 80.4% 80.0% 79.8% 78.3% 77.4% 76.7% 76.3% 76.0% 75.4% 74.9% 74.8% 73.7% 73.7% 72.3% 72.3% 70.5% 69.6% 69.5% 69.3% 67.0% 66.5% 66.2% 66.2% 65.3% 65.1% 63.1% 64.2% 64.1% 62.7% 61.1% 60.3% 60.1% 59.3% 59.1% 58.7% 57.8% 57.5% 57.1% 56.1% 56.0% 55.5% 55.1% 54.6% 53.4% 52.6% 51.6% 48.0% 47.8% 47.4% 47.2% 46.3% 45.9% 44.9% 42.3% 42.2% 42.1% 41.8% 41.5% 41.2% 40.5% 40.2% 40.2% 39.2% 37.8% 37.4% 36.3% 31.2% 30.4% 25.1% 25.0% 22.4% 15.0% 9.6% 7.0%

4 2 13 3 3 17 12 3 15 12 24 3 4 10 10 11 34 8 2 3 14 10 21 13 3 7 2 8 16 24 20 7 20 9 7 19 19 31 13 5 3 3 38 25 7 9 12 12 4 21 43 11 9 3 5 3 3 13 8 32 14 6 16 10 14 3 5 2 5 11 23 2 10 8 3 19 11 5 19 10 3 2 17 7 13 2 4 6 5 2 4 2

Artemis Unit Trust Managers Mayflower Management Britannic Fund Managers Sarasin Investment Funds Thesis Asset Management Framlington Investment Management Halifax JO Hambro Capital Management Legal & General ISIS Gartmore Investment Management Exeter Asset Management BWD Rensburg Rothschild Asset Management Dresdner RCM Global Investors Edinburgh Fund Managers Merrill Lynch Cazenove Fund Management Marlborough Fund Managers TU Fund Managers Aegon Group Scottish Mutual Old Mutual First State Capital Management Morgan Stanley Quilter Royal London Lloyd George Management Credit Suisse Fidelity Investments Mellon Financial Threadneedle Asset Management St James's Place Barings Canada Life Management Global Asset Management Aberdeen Asset Managers Standard Life Investments INVESCO Societe Generale Sovereign Unit Trust Mgt. LeggMason Group CIS Schroders M&G Henderson Global Investors Axa Lazard Asset Management Govett Investment Management Family Assurance Friendly S. Hill Samuel JPMorgan Fleming Insight Investment Rathbones Abbey Unit Trust Managers Tilney Investment Management Taube Hodson Stonex Fuji Lord Abbett Martin Cume Management State Street Scottish Widows Jupiter Investec Asset Management Baillie Gifford NPI Investment Managers Smith & Williamson Franklin Templeton Pearl Unit Trusts Cavendish Unit Managers Prudential Corporation Capita Financial Managers Dresdner Kleinwort Marks & Spencer Deutsche Bank/DWS/Deutsche Lincoln Unit Trust Mgt. Liontrust Asset Management Scottish Amicable Abbey Nationat/Inscape Scottish Life HSBC F&C Hargreaves Lansdown Sand Aire Unit Trust Mgt. Friends Provident Barclays Singer & Friedlander Solus Investment Funds Bank of Ireland Asset Mgt. ABN Amro Investment Funds MGM Unit Managers Gerrard Investment Funds Allchurches Investment Mgt. Wesleyan Unit Trust Mgt.

7th April 2004

Stephen Wynn,
October 2006.