We are testing a new beta website for gov.scot go to new site

Scottish Economic Statistics 2008

Listen

A3 Financial Intermediation Services Indirectly Measured ( FISIM)

Stevan Croasdale and Andrew Mortimer, Scottish Government

Introduction

This paper seeks to explain the major methodological improvements that are being carried out in the calculation and allocation of Financial Intermediation Services Indirectly Measured ( FISIM). These changes are planned to come into effect with the publication of the Scottish Supply and Use tables time series data for 1998-2004, due to be published in early 2009; and the Scottish Quarterly GDP index for 2008 quarter 3, due to be published in January 2009.

Valuation of an industry's output and value added

Before exploring the particular challenges associated with valuing the contribution of financial institutions to the economy, it is perhaps useful to cover briefly how a 'typical' industry is valued in the National Accounts. The diagram shows the key measures taken from an aggregate 2004 Input-Output 'use' matrix.

Figure A3.1: Excerpt from Aggregate Combined Use Matrix 2004 (Purchasers' Prices) 1

Figure A3.1: Excerpt from Aggregate Combined Use Matrix 2004 (Purchasers' Prices)

Using agriculture, forestry and fishing as an example, we can see from "A" that its total output in 2004 was £3,535 million, that is, the total value of the goods and services produced by this industry. Item "B" shows the intermediate consumption of agriculture, forestry and fishing (£2,207 million), that is the goods and services purchased by the industry from other industries (and itself) in producing their output. The residual "C" (£1,328 million) is equal to output "A" less intermediate inputs "B" and is known as Gross Value Added ( GVA) - literally the value added to raw materials and other goods and services by the agriculture, etc industry.

The above calculations, in theory, could be repeated for every other industry to form a measure of whole economy GVA or Gross Domestic Product at Basic Prices. However, the process is complicated by the financial services sector due to the nature of its output. If we were to value financial services output "A" in the same way as shown above - very broadly as the value of their sales - then we would find that the financial services sales, despite the obvious size of the sector, would be very small. Once their intermediate inputs were subtracted, would leave an industry with a value of GVA which was exceptionally small, possibly even negative. However, "sales" doesn't accurately measure the scale and scope of the services provided by the financial services sector.

What is FISIM?

Financial institutions do make explicit charges for some services, such as commission on foreign exchange, account charges and flat rate fees for overdrafts. However, many of the services provided by financial institutions are ostensibly free: they often provide current account facilities to households free of charge, along with a large number of other services like financial advice.

Instead, and unlike any other industry, financial institutions rely extensively on revenues accruing from interest flows. This is the interest differential between the amounts paid by financial institutions to depositors and that charged to borrowers in the form of interest. To use a simple example: person "A" deposits a sum of money into their savings account and, over time, receives interest on these savings amounting to £1,000, the bank lends person "A"s money to person "B" who is charged a higher rate of interest, amounting to £2,000 over the same period. In doing so, the bank is providing a service to both persons even though there are no explicit charges to either person; the total value of the services provided by the bank is equal to the interest differential (£1,000).

In order to measure accurately the size of the financial services sector, National Accounts must incorporate an indirect measure of the value of the services provided where there is no explicit charge - this is FISIM - this indirect measure is added to the explicit charges to form a figure of total output for financial institutions.

The inherent problem associated with the incorporation of FISIM is imbalance - the known supply of banking services (direct + indirect output, reported by financial institutions) does not equal the known demand for the financial services in the economy (reported by businesses and final consumers). In fact, customers of financial institutions may not be aware that a transaction has taken place ( e.g. Person "A" did not pay an overt charge for the security provided by the bank for their savings) and, if they are aware, they are very unlikely to be able to place a figure on the value of the services provided to them - beyond those services which attract an overt charge.

This imbalance manifests itself in the form of artificially increasing the level of Gross Value Added (or GDP at basic prices). As shown above, GVA in its simplest form is calculated as the total value of goods and services produced, less intermediate purchases: the total value of goods and services required by firms to produce their output (raw materials, components, energy; and services, e.g. accountancy, advertising, banking, etc). The problem with FISIM is that it is earned by financial institutions, and contributes to their GVA, but there is no corresponding purchase of FISIM by other industries - which would, if enumerated, increase the other industries' intermediate purchases and, by association, reduce their GVA.

Given that, by definition, supply must equal demand within National Accounts supply/use balances, it is necessary to compensate for indirect output by imputing purchases of indirect financial products into the economy. There are two means to achieve this: a single "whole economy" adjustment; or to formally enumerate the amount of FISIM 'purchased' by each industry/sector in the economy and adjust each individually.

Approach 1 - The Financial Services Adjustment

The current Scottish Supply and Use tables and Quarterly GDP indices use the simplified approach, where (by convention) FISIM output (or Financial Services Adjustment ( FSA)) is not allocated between users, but is treated as absorbed by the intermediate consumption of a notional industry, purely as a balancing item used to reconcile supply and demand within the economy. No attempt is made to apportion any FISIM out to the specific industries or final demand sectors that use the services. The FISIM (or FSA) industry is shown as purchasing all FISIM output attributed to the financial services industry. The FSA also contains a measure of GVA of equal and opposite value to its purchases of FISIM, resulting in zero output. As a result the full estimate of FISIM is allocated to intermediate demand, reduces overall GVA by the same amount added to the financial services industry, resulting in no change in Gross Value Added ( GVA).

Figure A3.2 Simplified diagram showing the effect of incorporating approach 1 into a Use matrix 1

Figure A3.2 Simplified diagram showing the effect of incorporating approach 1 into a Use matrix

Although simple, the key limitation of the Financial Services Adjustment is that FISIM added to financial services and the compensatory effect of the FSA occurs only within intermediate demand (trade between industries). Clearly, the financial services sector also profits from its interactions with households and other final consumers - this is ignored by the FSA. Another, minor, limitation of the FSA approach is presentational: the proportion of each "real" industry to total GDP at basic prices sums to more than 100%.

Approach 2 - The allocation of FISIM across sectors

The approach recommended in the System of National Accounts 1993, and the related European System of Accounts 1995, involves a full allocation of the use of FISIM across the Supply and Use tables within intermediate consumption, final consumption expenditure or exports according to which sector incurs the expenditure, rather than assuming FISIM is a nominal industry within intermediate demand. As a result, one effect of the second approach is to increase the level of GVA.

The diagram below shows the effects of incorporating approach 2 in the Use matrix of the Scottish Input-Output tables. It should be noted that this is a simplified version of the actual steps involved in the estimation, and there are complexities, particularly with regard to interactions between financial institutions and non-market producers, e.g. central government.

Figure A3.3 Simplified diagram showing the effect of incorporating approach 2 into a Use matrix 1

Figure A3.3 Simplified diagram showing the effect of incorporating approach 2 into a Use matrix

Applying approach 2 in the UK National Accounts and Scottish GDP estimates

Within the UK National Accounts, the simplified nominal industry approach has traditionally been used. UK Regional Accounts provided an estimate of the size of the Financial Services Adjustment published alongside GVA for 32 broad industry groups for Scotland. In turn these have been used as constraining totals within the Scottish Supply and Use tables. The Scottish tables allow these estimates to be further split to 128 industry categories from which the detailed weights for the component series used in the compilation of the quarterly GDP series are derived.

In light of revised ESA95 guidelines and improvements in data availability from the Bank of England, the UK has incorporated approach 2 - the full FISIM adjustment - within their Supply and Use framework. As a result, detailed data are available on purchases of FISIM by each sector of the economy which allows the implementation of approach 2 in the Scottish Supply and Use tables also.

The impact of this new approach was to increase the absolute value of the UKGDP at current prices by around 1.5 percent. A similar increase for Scotland is likely to appear in the next round of Regional Accounts, due to be published on 12 December 2008.

Timing

The first UKGDP estimates based on approach 2 were published on the 30 September 2008 and related to Quarter 2 2008. The first Scottish GDP estimates based on approach 2, relating to Quarter 3 2008, are expected to be published on 28 January 2009.

Conclusion

The incorporation of the full FISIM adjustment - approach 2 - into Scotland's economic statistics represents a significant improvement in the measurement of the interactions between the financial services sector and other sectors. The approach acknowledges the full scope of the industry by recognising, for the first time, the value generated in the economy through financial services' interactions with final consumers, notably households, government and exports. Through incorporation of the full FISIM adjustment we will improve compliance with international National Accounts guidance and maintain, once again, comparability with economic statistics produced for the United Kingdom as a whole.