taxation of financial services under a value-added tax - CiteSeerX

National Tax Journal Vol 50 no. 1 (March 1997) pp. 89-112

TAXATION OF FINANCIAL SERVICES UNDER A VAT

TAXATION OF FINANCIAL SERVICES UNDER A VALUE-ADDED TAX: APPLYING THE CASH-FLOW APPROACH SATYA PODDAR ENGLISH *

*

& MORLEY

Abstract - Financial services have generally been exempt under valueadded tax (VAT ) systems due to the inability to identify the appropriate tax base, which is hidden in the financial margin. A cash-flow VAT approach represents one means of defining this tax base in a way that is compatible with credit-invoice VATs. This paper describes the use of a cash-flow VAT for a banking deposit/loan operation and indicates the practical difficulties blocking use of the full cash-flow method. It then describes a variant of the cash-flow system, referred to as the “Truncated Cash-Flow Method with Tax Calculation Account,” which is designed to overcome these difficulties.

quence, almost all the countries employing a VAT have opted to exempt financial services rendered to residents of their countries. Although the exemption approach has been widely adopted, there have been significant problems in its operation. The resulting tax cascading creates numerous economic distortions and adversely affects the competitive position of domestic institutions in international markets, while the definitions and allocations needed to operate the system have proven to be very complex. This paper presents an approach to applying a VAT to financial services that is designed to operationalize the cashflow method. The cash-flow method has been considered a conceptually sound method of applying a VAT to financial services with some attractive attributes in the context of credit-invoice VAT systems. However, the method has also been seen to exhibit several fundamental difficulties, which have kept it from being used except in a few limited situations. The approach suggested here is designed to overcome these fundamental difficulties, while

One of the most difficult areas in the operation of value-added tax (VAT) systems has been the treatment of financial services. A workable method of taxing such services has eluded tax authorities for decades. As a conse*Ernst & Young, Toronto, Ontario, Canada M5K 1J7.

89


NATIONAL TAX JOURNAL VOL. L NO. 1

retaining the positive aspects of the cash-flow approach. The solution identified appears promising. However, it requires considerable elaboration and testing to ensure that it is indeed practical. This paper outlines the approach as a basis for comments and further research.1

flow approach which heretofore have blocked its adoption in a general VAT system. It deals with the basic deposit/ loan activities at a conceptual level. However, the intention would be for other financial services also to be taxable under the system proposed, if they are not already part of the regular credit-invoice system (as financial services rendered for a fee or commission). Other financial service activities include life and property and casualty insurance; purchase, sale, and issuance of financial securities; brokerage and other agency services; advisory, management, and data processing services; and specialized financial products such as bullion and derivatives.

The paper concentrates on the key conceptual issues involved in the design of the cash-flow approach being developed. It does not attempt to assess the administration or compliance costs that the system would impose, nor does it cover all of the critical design issues that would arise.2 Considerable further work, some of which is underway, will be required to consider all such issues.

Financial Services in Banking Transactions

BACKGROUND The cash-flow method has received considerable attention in the literature, both as an alternative form of a VAT and as a replacement for the corporate income tax. The classic treatment of this form of tax is found in Meade (1978), in which the approach is largely discussed from the perspective of serving as a substitute for the corporate income tax.3 However, the approach discussed in this paper is a cash-flow VAT applicable to financial services, which is designed to function as a component of a broad-based VAT system applying to both financial services and nonfinancial goods and services.

Financial intermediaries create value by controlling the cost for carrying out transactions in financial markets and helping reduce the costs of transactions in real goods and services.4 Financial intermediaries are able to charge for their services in two ways: (1) explicit fees and commissions and (2) implicit charges in the form of margin. Examples of explicit fees and commissions include safe-deposit box rental charges, fees for certain trustee services, commissions charged by brokers on acquisitions and dispositions of securities, and charges for issuance of traveler’s checks. In principle, there is no conceptual difficulty involved in the application of a VAT to these explicit charges. In fact, many of the countries that have VATs do apply them to at least certain services for which specific fees or commissions are the typical method for

This paper sets out the nature of financial services in a deposit and loan banking transaction, provides examples of how a cash-flow system functions to tax such services appropriately, indicates why a cash-flow tax is considered to have particular merit as a means of extending a VAT to financial services, and discusses the problems in the cash-

90



charging for the services. Safe-deposit box rental charges, debt collection charges, and fees for advisory services are taxed in all countries with a VAT.

each component under a general consumption tax such as a VAT is provided in brackets. (1) The initial deposit creates a right to a future cash withdrawal and the future cash repayments contingent on this right. (This merely represents transfer of funds and should not be taxable.) (2) Pure interest payments. (This is a compensation for deferral of consumption from one period to the next and should not be taxable under a consumption tax. These interest payments are income to the recipient and should only be taxed as consumption if and when 2 consumed.) (3) Pure risk premiums representing the amounts charged to borrowers 2 to cover the risk of default. They are equal to the expected value of defaults, excluding any profit element or administrative cost of the financial institution. (The pure risk component should not be taxable, as it is only a redistribu-

However, the situation is quite different in the cases in which the charge for the value-added takes the form of an implicit charge included in the margin on the exchange of funds associated with a financial transaction. Figure 1 indicates the basic nature of banking intermediation involving a deposit and loan transaction. Depositors may be households, businesses, or other entities, as may the lenders. In the example, the value of banking services is shown as the spread between the interest received by savers ($70) and that paid by borrowers ($150). The value added (before deduction of purchased inputs) in the transaction is $80.5 Analysis of this type of transaction indicates that the financial flows involved are of the four types listed below. The appropriate treatment of

FIGURE 1.

Banking

1 Households

1

Business

Other

Loans

1,000

Interest Received

150

Deposits

1,000

Interest Paid

(70)

Value of Banking Services Households

Business

Other

91

80



tion of funds that is a form of wealth transfer among the participants in the underlying transactions.) (4) The compensation to the financial institution for the costs involved in accepting deposits and making loans, including the profit earned on behalf of shareholders. (The amount remaining after netting out capital costs is the value-added through financial intermediation of the financial institution and the appropriate tax base for a general consumption tax.)

the value of financial services provided by financial institutions, while allowing business customers (registrants for the VAT) to obtain input tax credits providing a recovery of VAT charged on the financial services they utilize.6 3 The essence of the cash-flow method is that it treats cash flows from financial transactions in the same manner as such flows from nonfinancial transactions. Cash inflows from financial transactions are treated as taxable sales (e.g., a bank would remit tax on a deposit), and cash outflows are treated as purchases of taxable inputs (e.g., a bank could claim an input tax credit on a deposit withdrawal). An exception to this relates to share transactions in a company’s own equity. An issue of equity is not considered to give rise to a taxable inflow of funds, and dividend payments do not give rise to input tax credits.7 To zero rate financial services rendered to nonresidents, transactions with nonresidents are ignored for purposes of the cashflow tax—inflows and outflows are neither taxable nor creditable. There is no change to the tax treatment of nonfinancial transactions by financial institutions, assuming that these are already taxable under the normal creditinvoice rules. However, all input tax credits related to commercial activity are now claimable, not just those related to nonfinancial supplies. Illustrative examples can be used to demonstrate that the results are the same as they would be if the financial institution were able to identify the value-added in each transaction, charge tax on it, and provide the appropriate invoice to allow business customers to claim input tax credits.

While only the final component (net of capital costs) represents value-added, which should be subject to tax, it is typical for the exchanges of funds and the margins between such exchanges to commingle several of the types of financial flows. In many common situations, it is impossible to identify the value of financial intermediation associated with specific transactions with the certainty required in a tax system, especially since there may often be cross-subsidization of transactions. It is this inability to identify and isolate the financial intermediation margin on particular transactions that is the fundamental difficulty in applying a credit-invoice system of VAT to financial services in which the charge for the service is in the form of margin. Even in cases in which the gross margin can be identified, it is not possible to separate it into the value of services provided to the depositor and the value of services provided to the borrower, which is necessary if input tax credits are to be properly allocated to business customers. Taxation of Financial Services with a Cash-Flow VAT

Illustrative examples are provided for three cases, all involving a banking transaction consisting of the receipt of a deposit, a loan, and a repayment of the

This section describes how a cash-flow system would operate to tax correctly 92



loan and the withdrawal of the deposit with the payment of interest in both cases. It is possible to provide similar illustrative examples for other types of transactions and different categories of customers with the basic results continuing to hold.8 The three cases, which are set out in summary form in the indicated figures, are as follows:

interest. This would be equivalent to the rate that the government would pay on its short-term debts. In the examples, this government rate of interest is 12 percent. With seven percent being paid on deposits, the consumer is considered to be receiving five percent in intermediation services associated with the deposit. The value of intermediation services provided to the borrower by the bank is three percent, the difference between the interest rate on the loan and the riskfree government rate. The interest rates shown are tax inclusive. The rate of the VAT is ten percent.

(A) consumer depositor, consumer borrower (Figure 3), (B) consumer depositor, business borrower (Figure 4), and (C) resident consumer depositor, nonresident borrower (Figure 5).

Figure 3 sets out the case in which both the depositor and the borrower are individuals not engaged in commercial activities. In this case, the VAT should be collected on the value of the financial service to both the depositor and the borrower, as both represent personal consumption. The example indicates that the cash-flow method does successfully tax this margin.

It is useful to use a common set of assumptions about interest rates, the value of services, and the rate of the VAT to facilitate comparisons of different cases. These are detailed in Figure 2. A basic assumption in the examples is that the government can earn a rate of return on its cash balances that is equal to the pure rate of

FIGURE 2.

Common Assumptions For Illustrative Examples

Deposit Interest

7%

Loan Interest

15%

Pure Rate of Interest = Interest earned by government

12%

Value of Services to Depositor

5%

Value of Services to Borrower

3%

Total Value of Financial Services

8%

VAT Rate

10%

93



FIGURE 3.

Illustrative Example

A: Consumer Depositor, Consumer Borrower Bank Inflows

Bank Outflows

Tax/ Credits

Period 1 Deposit Loan

100

Subtotal

100

–100

10 –10

–100

0

–100 –7

10 1.50 –10 –0.70

–107

0.80

Period 2 Loan Repayment Loan Interest Deposit Withdrawal Deposit Interest

100 15

Total

115

Total Value of Banking Services VAT

= =

8 0.80

No further tax adjustment at the consumer level

Under the cash-flow method, the bank would be subject to a tax of $10 on the cash inflow from the deposit but would be entitled to an input tax credit of $10 on the cash outflow associated with the loan. As these two cancel out, there would be no tax revenue for the government and no tax implications for the bank. When the transactions are unwound in period 2, the bank would again have offsetting tax liabilities and credits arising from repayment of the loan (giving rise to a liability as a cash inflow) and the withdrawal of the deposit (giving rise to a credit as a cash outflow). However, the tax of $1.50 (ten percent of $15) associated with the cash inflow would exceed the input tax credit

of $0.70 (ten percent of $7) obtained as a result of the cash outflow as interest on the deposit. The bank would be required to remit the balance of $0.80 to the government as a VAT in respect of the financial intermediation services provided. The tax of $0.80 is ten percent of the value of banking services as required. Figure 4 provides an example in which the depositor is still an individual but the borrower is now a business engaged in a commercial activity. If the cash-flow method is to yield the correct result, there should be a net tax levied on the service provided to the individual depositor who is a consumer of financial 94



FIGURE 4.


B: Consumer Depositor, Business Borrower Tax Payments by Bank Period 1 Period 2

0 0.80

See example A Tax Payments by business borrower

Borrowers Inflows

Business Borrower Outflows

Business Tax/ Credits

Period 1 Loan

100

10

Period 2 Loan Repayment Loan Interest

–100 –15

Subtotal

–10 –1.50 –11.50

Govt Revenues Period 1 Tax Interest earned @ 12% Period 2 Tax (0.8 – 11.5) Total

10 1.20 –10.70 0.50

Equals 10% of the value of Banking Services to Consumer Depositor

services, but no net tax applicable to the services provided to the borrower, because the financial service in this case represents a business input.

now have tax implications for the business borrower. In the first period, the cash inflow from the loan gives rise to a tax liability of $10 (ten percent of the cash inflow of $100). In the second period, the loan is repaid with interest and there is a cash outflow of $115, giving rise to an input tax credit of $11.50. The present value of this credit

Figure 4 indicates that the tax remissions by the bank are the same as those in Example A, namely, $0.80 in period 2. However, the loan and its repayment 95



FIGURE 5.


C: Resident Consumer Depositor, Nonresident Borrower Tax Payments by Bank

Bank Inflows

Bank Outflows

Tax/ Credits

–100 –100

10 0 10

–100 –7 –107

0 0 –10 –0.70 –10.70

Period 1 Resident Deposit Nonresident Loan Subtotal

100 100

Period 2 Nonresident Loan Repayment Nonresident Loan Interest Resident Deposit Withdrawal Resident Deposit Interest Subtotal

100 14.70

114.70

Govt Revenues Period 1 Tax Interest earned @ 12% Period 2 Tax Total

10 1.20 –10.70 0.50

Equals 10% of the value of Banking Services to Resident Consumer Depositor

is equal to the present value of the borrower’s earlier tax paid at the borrower’s interest rate of 15 percent. The business borrower thus does not experience any net tax on its use of banking services.

figure, this provides the government with $11.20 in the second period. Meanwhile, in the second period, the government provides a credit of $11.50 to the business borrower, while receiving tax of $0.80 from the bank (as in the first example). The second period thus leads to a cash outflow of $10.70 for the government. When this is netted against the $11.20 the government has on hand, the government is left with $0.50 in tax revenues. This is equal to a ten percent tax on the $5 of financial services provided to the individual depositor.

The revenue implications of these transactions for the government are summarized at the bottom of the figure. As was noted earlier, the government could invest the $10 tax remitted in the first period at the government interest rate of 12 percent. As shown in the 96



The example thus indicates that the cash-flow method provides for refunding of tax to business users of financial services, while assessing tax on consumption of these services by households.

financial margins include a payment in respect of the time value of money, which compensates savers for delaying consumption. This is an income payment and should not be part of the tax base for the VAT. A fundamental attribute of the cash-flow method is that it excludes the time value of money without explicitly requiring it to be identified. There is thus no requirement that margin be disaggregated into its components for the value-added to be taxed. Clearly, this is very attractive from an operational perspective, because it is very difficult to carry out such a disaggregation.

The final illustrative example set out in Figure 5 involves a domestic individual depositor with a nonresident borrower. The deposit by the resident leads to a VAT of $10 on the cash inflow, and the government earns $1.20 interest on this, giving it funds available of $11.20 in period 2. When the resident individual withdraws his deposit with interest ($107 in total), the bank claims a credit of $10.70. The government receives net VAT revenues of $0.50, which is equal to the VAT rate of ten percent applied to the financial service worth $5 to the individual. The foreign loan has absolutely no tax consequences, as the inflows and outflows associated with transactions with nonresidents are ignored. As a result of this, the bank would be able to provide the loan to the foreign resident at the tax-exclusive interest rate of 14.7 percent, placing the bank in a competitive position with banks in other jurisdictions that may not be subject to consumption taxes with respect to financial services.9

Full removal of tax cascading

The cash-flow method allows for the full removal of tax cascading on financial services provided to registered persons. Registrants are able to claim input tax credits in respect of their cash outflows to financial institutions, while they are subject to tax on their inflows. In the case of nonfinancial enterprises, input tax credits on outflows will normally exceed tax on inflows. This excess compensates them for the tax they will be paying on their purchases of financial services. As was indicated by the examples, the operation of the system is such that the registrants are placed in identical positions to the ones they would be in if the tax were not in place.

Desirable Features of the Cash-Flow Tax Approach for Financial Services

Destination basis

The cash-flow system functions on a destination basis and provides for the equivalent of zero rating of services provided to nonresidents. Transactions with nonresidents do not lead to either the application of tax or the claiming of credits. Domestic financial institutions are not at a disadvantage in the supply of services to nonresidents. Because the cash-flow method is on a destination

The cash-flow method as outlined above exhibits a number of very desirable characteristics as a method of extending the VAT to financial services. Time value of money excluded with no explicit adjustment

In discussing the nature of value-added in financial services, it was noted that 97



basis, it is compatible with this aspect of the VAT for a credit-invoice system applying to other goods and services.

principal and interest upon repayment of the loan (equal to $11.50 per $100 of principal using the values from the earlier examples), despite never having paid tax on the cash inflow. This would clearly be quite inappropriate and lead to net debtors being overcompensated for the VAT on the financial services consumed.

Fully compatible with the credit-invoice method

In fact, the cash-flow method is fully compatible with the credit-invoice system nonfinancial services. As well as operating on the destination basis, it provides for the claiming of input tax credits in respect of financial services and thus eliminates cascading. Because value-added on financial services is taxed on an equivalent basis to other activities, it eliminates the non-neutralities which create adverse economic effects.

In theory, one could implement a system that required all registered persons to pay an additional tax equal to the tax rate increase times the excess of financial liabilities over financial assets just prior to the tax rate change. As a consequence, value-added earned prior to the change is taxed at the old rate and value-added earned subsequently is taxed at the new rate. While this approach is conceptually valid, it would be subject to political objections and would clearly raise cash flow issues for business.

It is for these reasons that the cash-flow approach has often been seen as the most promising mechanism for extending the VAT to financial services.

Second, any time registrants took out loans, they would be required to pay tax on the cash inflow. As a result, the borrowing requirements would increase in order to finance the tax payment associated with the loan. While the business would subsequently recover this tax at the time the loan was repaid, there would clearly be cash-flow implications that would be seen as problematic by nonfinancial businesses.

Difficulties with the Cash-Flow Approach Despite the attractive features of a cashflow approach to taxing financial services, there have been a number of fundamental difficulties identified that have kept this method from being adopted in practice.10 The first of these relates to the implications of tax rate changes, the most significant of which would be the rate increase associated with introducing the tax. This is a concern, because for the cash-flow method to function appropriately, the credit for the cash outflow should be at the same rate as the tax applicable to the tax inflow. A simple example demonstrates the issue raised. For example, suppose a VAT was introduced in which financial services had previously been exempt. If the system simply came into effect, a borrower would receive a credit for the

Finally, the cash-flow method, as outlined above, would require nonfinancial businesses to carry out all the calculations for the tax in order to obtain access to input tax credits for the financial services purchased. For smalland medium-sized businesses, this would be a considerable compliance burden and would considerably reduce the effective benefit from having access to the input tax credits. The system as implied in the earlier description thus 98



can be criticized as unduly complex for many businesses.

are subject to interest charges at the government borrowing rate.

CASH-FLOW METHOD WITH TAX CALCULATION ACCOUNT

Basic Features of the TCA The basic features of the cash-flow method with a TCA can be briefly summarized as follows:

This section describes refinements to the cash-flow system designed to deal with the problems identified in the preceding section, while retaining the general advantages of the cash-flow approach. The methods described revolve around a mechanism that is referred to as the tax calculation account (TCA). A second refinement involves “truncating” the system, by placing the major compliance activities solely with financial institutions.

(1) tax payments on cash inflows related to a financial instrument (whether an asset or a liability) debited to the TCA; (2) input tax credits on cash outflows related to a financial instrument credited to the TCA; (3) net balance in the TCA subject to an indexing adjustment at the government borrowing rate (as will be discussed later, a short-term government borrowing rate is a proxy for the pure rate of interest); and (4) a balance in the TCA payable (or refundable, if a negative amount) periodically, after subtracting a notional amount equal to the tax rate times the value of the financial instrument at the end of the period.

Cash-Flow Method with the TCA Two of the main difficulties identified with respect to the basic cash-flow system were cash-flow problems related to payment of tax at the time of borrowing and transitional adjustments at the startup of the system or at the time of a tax rate change. In each of these cases, the difficulty arises only with respect to margin services involving cash inflows and outflows of a capital nature. The treatment of fees and commissions and cash flows of a current nature related to margin activities do not lead to any problems.

Figures 6 and 7 illustrate the operation of the TCA for a $100 loan and a deposit transaction, respectively, assuming that the interest and tax rate values are the same as those used in the earlier illustrations. Under the full cashflow method, both the bank and the business registrant would have TCA accounts. It is assumed, in this example, that the deposit and loan occur at the end of period 1 and that there are no current interest amounts in that period. The deposit and loan are both assumed to be repaid at the end of the second period.

The TCA is a tax suspense account created to obviate the payment of tax by taxpayers and of credits by government during the period that cash inflows and outflows of a capital nature occur. Tax that would otherwise be payable/creditable is instead debited/ credited to the TCA and carried forward to the period during which the capital transaction is reversed. The TCA mechanism thus allows deferral of tax on cash inflows and of tax credits on cash outflows. However, these deferrals

In Figure 6, the bank has made the $100 loan, leading to a cash outflow and thus 99



FIGURE 6.

Cash Flow Method with TCA

Illustration TCA for loans by a bank

Loan

Amount

TCA

(100)

(10)

TCA Indexing

—

(1.2)

Interest

15

1.5

Loan Repayment

100

10

Loan Closing Value

—

—

Net Tax Due

—

FIGURE 7.

0.3


Illustration TCA for deposits with a bank

Deposit

Amount

TCA

100

10

TCA Indexing

—

Interest

(7)

Withdrawal

(100)

1.2 (0.7) (10)

Deposit Closing Value

—

—

Net Tax Due

—

0.5

100



a credit balance of $10 in the TCA. There will be a negative $1.20 indexing adjustment added at the end of the period. The repayment of the loan will lead to a cash inflow of $100 and a debit of $10 in the TCA. The receipt of loan interest of $15 will result in a debit entry of $1.50. The balance of these entries is a positive amount of $0.30, which must be paid to the government as tax.

ensures that any tax or credits actually paid are not related to capital amounts. The notional adjustment as defined earlier is carried out by subtracting a notional amount equal to the tax rate times the value of the financial instrument at the end of the period. This calculation is, in effect, equivalent to simply moving the loan repayment entries one line lower in Figure 6, to be shown as a loan closing value, and the deposit withdrawal entries one line lower in Figure 7, to be shown as the deposit closing value. The tax and credit calculated for the two accounts are then identical to those previously described, where the accounts are closed by a loan repayment and a deposit withdrawal.

In Figure 7, the operation of the TCA for the $100 deposit is described. The cash inflow of $100 gives rise to a debit of $10 in the TCA. Indexing at 12 percent for the period results in a further debit entry of $1.20. The withdrawal of $100 as a cash outflow gives rise to a credit of $10, while the interest payment results in a further credit of $0.70 in the TCA. The balance after all of these entries is a debit balance of $0.50, which is payable to the government as tax.

The TCAs in the books of a business borrower or a business depositor are the mirror images of the TCAs for a loan or a deposit in the books of the bank. For example, a business firm with a $100 loan from a bank will have an initial debit entry of $10 (at a tax rate of 10 percent) in the TCA, to which indexing of $1.20 will be added at the end of the year. Payment of interest of $15 on the loan in the year will result in a credit of $1.50. Finally, repayment of the loan will result in a credit of $10 for the cash outflow associated with the repayment. The overall balance will be a negative amount of $0.30, which will be refundable to the business as an input tax credit. This is the mirror image of the $0.30 tax calculated by the bank in its TCA for the loan to the business. This example shows that the net tax payable of the bank TCA becomes the tax creditable per the business borrower TCA.11

It can be noted that the total tax payable by the bank is $0.80 in respect of both transactions, the same amount as that calculated for a complete deposit/loan transaction in the earlier examples using the basic cash-flow approach. The figures can also be used to indicate the implications of a situation in which the interest payments are made in the period, but the loan and deposit principal amounts are not repaid. In that case, there is no entry for the loan repayment or deposit withdrawal in the two figures and no associated debit and credit entries in the TCA. However, because the accounts are still open since they contain an asset and a liability, respectively, a calculation of the notional value of the debit/credit that would be associated with closing the account needs to be made, as described in the summary of the system above. This

As can be seen from the examples, the TCA system eliminates any cash-flow problems by deferring tax payments and credits on capital transfers. For example, the commercial borrower is no 101



longer required to remit tax because of the inflow of funds from the loan in period 1. All that happens at the time of borrowing is that an amount is credited to the tax calculation account. Despite the fact that the government does not receive the tax in respect of inflows of cash related to borrowings, it is at little risk in revenue terms. The expectation is that the amounts that would otherwise be remitted will be reversed by a tax credit for the capital outflows associated with loan repayments and interest.12

(1) It eliminates cash-flow problems by deferring tax payments and credits on capital transfers. (2) Government is at little risk in respect of tax deferrals, because tax payments on capital inflows (e.g., borrowings) are expected to be reversed by tax credits for capital outflows (e.g., loan repayments), except in the case of bankruptcy. (3) The benefit/loss from deferral of tax or credit is offset by the indexing of outstanding TCA balances by the government borrowing rate. (4) Transition difficulties are addressed through initial debiting/crediting of the TCA at the commencement of the system and adjustment of the outstanding balance at the time of any tax rate change.

The TCA balances can be harnessed to deal with tax rate changes. The outstanding TCA balances would be grossed up at the time of a tax rate increase so that the tax/credit in respect of capital flows before or after the change is the same as the tax/credit adjustment in respect of the reverse flows after the change. Similarly, a tax rate reduction would be handled by grossing down TCA balances at the time of the rate change. As a result, there would be no immediate cash-flow consequences from the rate change for either government or businesses, and tax/credits would be correctly calculated at the new rates subsequent to the rate change.

Tax Calculation Account: Illustrative Example of System Implementation Figure 8 provides an example of the functioning of the TCA during implementation of the system. It is assumed that the system is introduced at the midpoint of a year. If the system is to function correctly, only one-half the value-added at the bank should be taxed for either a loan or deposit transaction. Only the case of a loan is shown, but it is easily surmised that similar results would occur for a deposit transaction.

The introduction of the system would proceed in the same manner. A debit/ credit balance would be created in the TCA at the start of the system, equal to the tax rate times the value of a given financial liability or asset. There would be no cash flow consequences of this opening adjustment for business at the startup, and subsequent tax/credits would then be determined under the general procedures described above.

At the bank, there were no tax implications when the original loan was made. At the point the system is introduced, the current value of the loan will be ascertained, including accrued interest. The bank will create a TCA entry equal to the tax rate times the value of financial assets as the starting value as of the date of implementation. In order for the system to exclude fully valueadded prior to implementation of the

The characteristics and advantages of the system relative to the basic cashflow system can be briefly summarized as follows. 102



FIGURE 8.


Illustration TCA for loans by a bank

Amount

TCA

Loan

(100.00)

(10.00)

Opening Balance

(107.24)

(10.72)

TCA Indexing

—

(0.63)

Interest

15.00

1.50

100.00

10.00

Loan Repayment Loan Closing Value

—

—

Net Tax Due

—

0.15

system, the assets (and liabilities) that are used to calculate the TCA at implementation must include amounts for the accrued interest on the deposit and loan.

in fact, equal to one-half year of valueadded in respect of the loan portion of the bank’s financial activities. A similar example could be presented for a rate change. A rate change would require closing an existing TCA just before the rate change and restarting it just after the rate change. In this case, the opening value of the TCA balance (which is the closing value of the closing TCA balance at the end of the previous period) would be grossed up to reflect the extent of the rate increase.

In the example, the negative TCA in respect of the loan would be $10.72 (the tax rate of 10 percent times the principle of $100 compounded for onehalf of a year at 7.24 percent, which is equivalent to an annual rate of 15 percent). The TCA indexing is then applied for the remainder of the year at the indexing rate of 12 percent annually. The value of such indexing compounded to the redemption point of the loan six months later is $0.63. When the loan is repaid, the $100 loan repayment creates the usual $10 credit entry, and the interest cash flow adds a credit of $1.50. The tax payable is $0.15. This is,

The Tax Calculation Account: Issues To ensure that the tax calculation account functions correctly, there are a number of issues that must be addressed. These include the choice of the 103



indexing rate, the treatment of the indexing rate outside the deposit/loan interest range, the rate changes in the indexing rate, the frequency of indexing adjustments, the asset valuations for system implementation or rate changes, and the compliance concerns.

deposit and lending rates. For example, the indexing rate could be 12 percent, but a particular institution could have a deposit rate of 14 percent and a lending rate of 22 percent, where the rates are locked-in for a period of, say, five years. This pattern could occur because of the term structure of interest rates and the institution-matching deposits and loans for a given maturity period. For example, the short-term interest rates could be in the 12 percent range, but the five-year interest rates could be much higher because of the traditional upward sloping yield curve. A locked-in interest rate for a five-year loan implies additional market risks for the lenders, but these risks can be eliminated for the bank by matching the maturities for the deposits and loans. The floating rates could still be 7 percent for deposits and 15 percent for loans as in other examples.

Choice of the indexing rate

The indexing rate in the cash-flow method with the TCA plays the critical role of dividing the margin between borrowers and lenders. It thus determines the size of the input tax credit that will be available for the registered users of financial services. It also serves the purpose of charging/crediting interest on the outstanding tax balances. As the rate that charges/credits interest on outstanding tax balances, it should be the interest rate that the government would have earned/paid on the tax/ credit amounts had there been no deferral. From the perspective of dividing the margins, the indexing rate should be a pure rate of interest that does not include any margin for financial intermediation or risk premium. Conceptually, any variation between the short-term and long-term interest rates reflects risks (bad debt risks or market risk of change in interest rates) or value of intermediation services.

The 10 percent difference between the 12 percent indexing rate and the 22 percent lending rate could be viewed as consisting of two components: a 3 percent charge for the normal banking services and an additional 7 percent charge for the market risks inherent in a five-year locked-in investment. By the same token, the negative two percent difference between the indexing and the deposit rates could be viewed as consisting of a five percent service charge by the bank to the depositor, and a reverse seven percent charge by the depositor to the bank for the additional market risk the depositor is taking by making a locked-in deposit for five years.

From both of these perspectives, it appears that a short-term treasury bill rate would be a good approximation of the cost of short-term money for government and a pure rate of interest that does not contain any risk or intermediation elements.13

Conceptually, both the service charge and the market risk premium are a consideration for financial services and should be included in the tax base. If the bank, the depositor, and the

Indexing rate initially outside range of the deposit and lending rates

In some cases, the indexing rate could initially be outside the range of the 104



borrower were all registrants, then the use of the 12 percent indexing rate by all three would yield an appropriate outcome. The depositor would remit tax on two percent, the bank would remit tax on eight percent (ten percent service charge to the borrower, offset by two percent net payment to the depositor), and the borrower would get credit with respect to ten percent service and risk charge to the bank.

(2) new rate lower than the deposit rate, and (3) new rate higher than the lending rate. The first situation involves a change of the indexing rate, but the new rate is still between the deposit and lending rate. In this case, there would simply be a change in the allocation of the margin between the depositor and the borrower.

However, if the depositor is not a registrant, then the bank should not be allowed to claim a deduction for the net service payment of two percent to the depositor. In this example, the depositor is a net supplier of services to the bank. The services supplied are protection against the risk of movement in deposit interest rates over the five-year period. The service provided by the unregistered depositor is to be treated as any other supply by a small supplier. It is neither subject to tax nor eligible for a credit.

If the rate goes down, the margin allocable to the depositor goes down and that to the borrower goes up. As a result, the lender would be a able to claim a larger input tax credit. This result is equivalent to full accrual of all capital gains and losses. When the indexing interest rate goes down, with the deposit and lending interest rates remaining fixed, the depositor enjoys a capital gain on the deposit that is locked-in at the old, higher interest rate, and the lender suffers a capital loss. The capital gain enjoyed by the depositor could be viewed as reducing the net payment to the bank for financial services. On the other hand, the loss to the borrower could be viewed as an increase in the payment to the bank for financial services.

These are clearly issues on which further input from financial institutions and other financial experts are essential. They need to be reviewed further from both conceptual and administrative perspectives. Changes in the indexing rate

Example: Initial lending rate 15 percent and indexing rate 12 percent imply a service charge of 3 percent to the borrower. If market conditions change and the indexing rate falls to ten percent, the lender suffers a capital loss of two percent for each year the borrowing rate is lockedin. If the borrower were to pay down the old loan immediately and refinance it at the rate of 13 percent, the bank would charge a penalty of 2 percent per annum on top of the principal amount of the loan outstanding. This additional penalty would be viewed as interest or an additional service charge and would be subject to tax in the hands of the bank. The

Short-term interest rates change over time, and, thus, the indexing rate is subject to fluctuations. Since deposits and loans may have fixed interest rates over a period of years, changes in the indexing rate would result in modifications to the margin assigned to the borrowers or lenders. A number of conditions could occur, and some of these appear on the surface to lead to rather unusual results. The possibilities are as follows: (1) new rate still between the deposit and lending rates, 105



borrower would receive an input tax credit in respect of this additional service charge and an ongoing credit for the regular service charge of three percent. If the loan is not paid down, then the use of the ten percent indexing rate gives exactly the same result, i.e., credit for a total service charge of five percent.

intervals, e.g., only at the beginning of a calendar month. It is largely an administrative and compliance matter, as there are certain considerations that limit the potential for tax planning based upon infrequent indexing adjustments. From the government’s point of view, the indexing adjustment may not imply significant revenue loss because it merely serves to divide bank margins between borrowers and depositors, leaving the overall taxable margin unaffected. Of course, the division would have indirect revenue consequences if the borrowers were businesses who claimed an input tax credit and the depositors were consumers who could not claim any credit. In the case of bearer transactions, an increase in indexing rate would reduce the tax base for financial assets, but would increase it for liabilities. In reality, given that both depositors and borrowers would include a combination of business and non-business customers, and given that financial institutions would have both bearer financial assets and liabilities, the risk of inappropriate tax planning behavior on the part of financial institutions from infrequent indexing rate adjustments is likely to be small as long as interest rate volatility is limited.

Therefore, the indexing mechanism has a very desirable property of automatically giving recognition to changes taking place in the implicit service charge due to market conditions. Frequency of indexing adjustments

For the system to yield the absolutely correct result, the indexing adjustment needs to be applied on a compound basis to whatever the outstanding balance is each day over the course of the tax period. This provides a measure of the accrual of indexing adjustment in a manner akin to the accrual of interest. This reflects the fact that the indexing adjustment is, in fact, a mechanism for charging/crediting interest on deferred tax balances. To achieve the appropriate results, the frequency with which indexing adjustments would need to be calculated would vary greatly from asset/liability to asset/liability. For fixed term loans, calculations once a year might be sufficient, while for deposit accounts, daily compounding would be appropriate (subject to possible de minimus rules). In practice, the legislation could probably give institutions considerable flexibility in the choice of the frequency that they consider appropriate, as long as it is calculated at the same annual compound rate.

Valuation of assets and liabilities at the commencement of the system or for a change of tax rate

A valuation of financial assets and liabilities will be needed at the time of introduction of the cash-flow VAT for financial services and at the time of any rate change. The valuation at commencement is necessary to restrict the application of the VAT and the availability of credits to postcommencement services. The valuation for rate changes is to ensure that accrued service values at the time of the change are subject to

Where interest rates change frequently, a decision has to be made whether the indexing rate would be changed at the same time or only at fixed periodic 106



tax and credit at the old rate and that subsequent services are subject to the new rate.

Under the truncated system, for loans and deposits with a financial institution, the TCA calculations could be done by the financial institution, with a periodic statement issued by the institution for the net tax credit claimable by business customers. With the exception of the indexing adjustment, the TCA entries would equal the tax rate multiplied by the entries in the deposit or loan account. The resulting statement issued to the customer by the financial institution on the basis of TCA calculations becomes a tax invoice and serves the same purpose as a tax invoice for nonfinancial goods and services. For a truncated cash-flow approach to apply to all types of financial services, a definition of financial institutions along the lines described in the previous section would be required to determine which taxpayers should have to carry out the TCA calculation and issue invoices.

The valuation is required for financial assets and liabilities only. Nonfinancial properties do not require valuation. Ideally, the adjustments in the TCA should be based on the fair market value of assets and liabilities at the time of change. The values must include accrued, but unpaid interest. This would ensure that the tax or a change in its rate did not have any retroactive application. In the case of loans, any loans that had become bad or doubtful before the commencement of the system would be discounted and valued at their realizable value. Otherwise, the system would result in the bank obtaining a tax credit for bad or doubtful debts that had occurred before the commencement of the system but that were not written off in the books until after the commencement of the system.

Conceptually, based on the above discussion, one can consider that there are three alternatives for a cash-flow approach when applying a VAT to deposit and loan banking services. These are as follows:

Truncated Cash-Flow Method with a TCA It was noted earlier that the cash-flow transactions of the financial institution form a mirror image of the cash-flow transactions of the business registrants in respect of deposit and loan transactions. This observation can be harnessed to provide one further system proposal that responds to a fundamental difficulty with the cash-flow approach. In this case, it is the complexity that would be experienced by smaller businesses in carrying out the cash-flow computations. The proposal would involve all the cash-flow calculations being undertaken by the deposit and loan intermediaries, thus truncating the calculations required under the general cash-flow approach.

(1) the general cash-flow system in which all registrants calculate taxes and credits on a cash-flow basis; (2) the cash-flow tax with a TCA in which all registrants calculate tax and credits using the tax calculation account; and (3) the truncated cash-flow system with a TCA in which financial institutions carry out the TCA calculations and nonfinancial registrants have access to input tax credits based on invoices supplied by financial institutions. In most respects, the truncated cashflow system with the TCA gives almost 107



identical results to the full cash-flow approach. It also provides a response to each of the fundamental difficulties that have been considered to exist in respect to the cash-flow approach and, thus, holds out considerable promise as a feasible option. However, further design work, testing, and analysis are necessary before such a system can be implemented.

Taxable persons would not include consumers and individual investors nor mutual funds and other pooled investment vehicles making portfolio investments on behalf of fund members. The rationale set out for the exclusion of portfolio investments by individuals applies in toto to the investment activities of persons not engaged in any other commercial activity, and they thus should not be taxable persons for purposes of the cash-flow tax. The mutual fund exemption merely extends the personal level treatment to pooled investment vehicles.

Definition of Financial Institutions The definition of financial institutions would obviously be important under the truncated cash-flow system, because only financial institutions would carry out the cash-flow calculations and would have the responsibility for issuing invoices to business customers indicating the amount of tax paid with respect to their financial services. Prior to discussing the definition of financial institutions in the truncated system, it is useful to note certain considerations that would arise in defining taxable persons under a full cash-flow system for financial services.

It would be expected that most enterprises, other than those acting as financial intermediaries, would realize a tax saving by being designated a taxable person. In other words, by being taxable, they would be able to receive input tax credit in respect of their financial transactions with taxable financial institutions. This would especially be the case when they had the option to exclude portfolio investments. Enterprises other than financial institutions that are engaged in the financial intermediation business (e.g., credit card services provided by large vendors) would be required to be registered for purposes of the cash-flow tax. This would ensure a uniform treatment of all enterprises providing such services.

The full cash-flow method would involve all persons involved in carrying on a commercial activity applying the cash-flow computations to their financial transactions, unless a specific exception applied.14 Taxable persons thus would include the following:

Under the truncated version of the tax, only the first two of the three types of taxable persons identified would be considered financial institutions and have a responsibility to do the cash-flow calculations and issue invoices to customers in regard to the tax on services. Other businesses with financial activities would not need any special cash-flow calculations, but would be able to claim input tax credits in respect of the

(1) financial institutions; (2) other businesses acting as financial intermediaries, e.g., a retail store with credit card operations; and (3) other businesses with financial transactions linked to their nonfinancial commercial activities, e.g., a manufacturer borrowing or lending funds in the course of its manufacturing activities. 108



VAT on financial services on the basis of invoices provided by financial institutions.

are created by ignoring the financial services of nonfinancial institutions. Distortions arise only when businesses excluded from the definition of financial institutions provide substantial margin services to consumers and other nonregistered persons. A practical review of the extent of such situations would be necessary to determine if, and where, such a supplementary category of financial business exists. Such businesses could be required to segregate such operations and to follow the rules applicable to financial institutions. In that event, the definition of financial institutions becomes redundant. What is needed then is a definition of financial business, regardless of whether it is carried out by a financial institution or a nonfinancial enterprise.

The definition of financial institutions could include the following: (1) banks, credit unions, and trust, loan, and acceptance companies; (2) credit card companies; (3) Investment dealers; (4) life insurers and property and casualty insurers; and (5) any other person whose principal business is lending money, accepting deposits, or purchasing or selling debt securities. Pension funds, investment funds, and financial holding companies, which provide financial services primarily to their affiliates, could be excluded from the definition. Their activities tend either to be in the nature of portfolio investments (as opposed to intermediation activities between depositors/savers and borrowers/users of funds), which would be excluded from the tax base were the investments to be held directly by individuals, or of services to other registrants, who could claim full credit in any case for the tax charged on financial services.

Other Design Features Implementation of a cash-flow variant of the VAT for financial services would necessitate dealing with several specific tax design features in order for the tax to function appropriately. The main examples of such design features relate to the tax base, exclusions from the tax base, treatment of bad debts, transactions with shareholders/owners, transactions between financial enterprises and nonfinancial persons, portfolio investments, definition of taxable persons, secondary market transactions, transactions with nonresidents, place of supply, definition of financial institutions, and financial services rendered by nonresidents. The design of the tax would also need to be extended to other types of financial services such as life and property and casualty insurance, investment dealer services, and credit card services. Some of these areas are the subject of ongoing research.

Consideration would have to be given to including in the cash-flow base the financial transactions of any other person who carries on a financial business on a regular and continuous basis, even if such activities are not the principal business of the person. An example of this would be credit card services to customers provided by department stores. The deciding factor in whether to include such activities in the cash-flow system would be a pragmatic one, depending upon the extent of competitive distortions that 109



approach during its development. Comments from the referees provided valuable assistance in improving the presentation and in identifying issues for further consideration. 2 The issues for which further design work are necessary are identified toward the end of the paper. One crucial design issue considered in the paper but that requires further analysis is the selection of a so-called indexing factor. The indexing factor is the interest rate that is used to allocate the financial margin between suppliers of funds and users of funds under the cash-flow method approach proposed. 3 Other studies that discuss the tax with particular emphasis on its applicability to financial services include Barham, Poddar, and Whalley (1987); Hoffman, Poddar, and Whalley (1987); and Hoffman, (1988). 4 A sample of studies that consider various aspects of the value-added by financial transactions includes Meade (1978); Barham, Poddar, and Whalley (1987); Hoffman, Poddar, and Whalley (1987); Hoffman (1988); and Henderson (1988). 5 The value-added would be disbursed as payments of salary and wages to labor, payments of dividends to owners, and purchase of other inputs used in the financial intermediation activity (for example, computers and rental payments for real property). There might also be a portion used to offset losses on bad debts. 6 The examples provided of the operation of a cashflow tax reflect material contained in a discussion of a cash-flow VAT contained in Poddar and English (1994). 7 The appropriate treatment of equity transactions is discussed in Meade (1978). Essentially, if equity investments were treated in the same manner as other financial transactions, economic rents earned by shareholders would be sheltered from taxation in addition to the normal rate of return reflecting the time value of money. 8 For clarity of exposition, the examples assume that the term and value of the deposit equal the term and value of the loan. However, the cash-flow method essentially operates on each customer’s transactions independently and provides the appropriate result where, for example, a series or group of deposits funds a loan. 9 Current VAT systems provide for zero rating of financial services that are exported. While determining the place of supply and allocating transactions among exempt and zero-rated activities are a significant source of complexity, conceptually, zero rating does eliminate any bias against domestic institutions on exported financial services. However, financial services supplied by offshore financial institutions into the domestic market do have a tax advantage over those supplied by domestic institutions. Foreign institutions do not pay any tax on inputs, either because of zero rating or the

While much remains to be done, our preliminary examination suggests that the cash-flow system can be extended to most financial services in a satisfactory manner. There would undoubtedly be a need for special rules and compromises to deal with specific sectors and transactions. These features would also be an element in the potential overall administrative and compliance burden of the system. Conclusions This paper has outlined a mechanism for applying a cash-flow system of a VAT to financial services. The principal design innovation is the development of a tax suspense account referred to as the TCA. Means of limiting the necessary tax calculations to financial institutions have also been identified. These approaches deal with the fundamental difficulties relating to system startup, tax rate changes, cash flow impacts, and complexity, which have generally been considered to stand in the way of implementing this approach to applying a VAT to financial services. The complete mechanism proposed is referred to as a truncated cash-flow method with a TCA. ENDNOTES 1

The mechanisms discussed in this paper were developed during research projects carried out for the Commission of the European Communities. However, the views expressed are those of the authors and not of any of the organizations providing research support for the project. The authors wish to thank Michel Aujean, Director, Indirect Taxation, and members of the VAT and Other Turnover Tax Unit, European Commission, Directorate-Génerale XXI, for their assistance and comments in this research. David Leslie of Ernst & Young, Toronto, played a critical role in the early development of the system. The authors also wish to thank Richard Bird, Glen Cronkwright, Jack Mintz, and Bill Vandeburgh for their detailed comments, as well as the numerous other individuals who have provided comments on the 110



absence of a VAT in the foreign jurisdiction, while the domestic institutions do bear (noncreditable) a VAT on their inputs. Extension of a VAT to financial services thus would eliminate this adverse economic effect for domestic financial institutions in Europe. 10 New Zealand has applied a cash-flow system to property and casualty insurance services. However, the services provided to borrowers from the insurers have not been taxed, and the system is a partial one in that respect. 11 It is useful to observe at this point that it is this mirror image characteristic that is used in developing a truncated version of the system. Under that system, all the calculations would be done by the financial institutions, which could carry separate TCAs for each customer. This would not only identify the tax payable by the bank in respect of the financial services provided to the customer, but would also provide the means by which the bank could provide an invoice for taxes paid, which would allow a business customer to claim input tax credits. 12 One situation in which there would not be an offsetting entry would be as a result of bankruptcy. Design of a cash-flow system would require development of appropriate rules for the treatment of bad debts. 13 While the initial view in developing the TCA system was that a single interest rate, excluding all risk and intermediation services, should be selected as the indexing rate, with a short-term government borrowing rate appearing to approximate such a benchmark, further consideration needs to be given to this issue. Among the specific questions to be addressed would be the following: (1) Is a single rate more appropriate than a cluster of rates reflecting different maturities of loans and deposits? (2) If a single rate is used, what is the best proxy? (3) Given that no interest rate may provide a perfect proxy, what are the nature and magnitude of any distortions? Further research is being undertaken on this issue. 14 A major issue exists with respect to the appropriate treatment of portfolio investments by nonfinancial registrants. While a strong case can be made for excluding transactions with respect to portfolio investments as a parallel to the treatment of individual portfolio investments, this would raise concerns where treasury operations are actually a rather significant factor in the overall activities of a nonfinancial commercial undertaking. This important issue is not discussed in this paper.

Broad-Based Value-Added Tax for the United States.” Tax Lawyer 28 (1975): 193–209. American Bar Association Section of Taxation. “Evaluation of an Additive-Method Value-Added Tax for Use in the United States.” 1977 ABA Section Tax Report: The Special Committee on the Value-Added Tax. Tax Law 30 (1977). American Bar Association Section of Taxation. “Value Added Tax: A Model Statute and Commentary.” Report to Committee on Value Added Tax of the American Bar Association Section of Taxation, Washington, D.C., 1989. Barham, Vicki, Satya N. Poddar, and John Whalley. “The Tax Treatment of Insurance under a Consumption Type, Destination Basis VAT.” National Tax Journal 40 No. 2 (June, 1987): 171–82. Canadian Department of Finance. Tax Reform 1987: Sales Tax Reform. Ottawa, 1987. Clarkson Gordon (Ernst & Young). Taxation of Financial Institutions, An Analysis of Canadian Tax Reform Proposal. Toronto, 1987. Henderson, Yolanda K. “Financial Intermediaries Under Value-Added Taxation.” New England Economic Review (July–August, 1988): 37–50. Hoffman, Lorey Arthur. “The Application of a Value-Added Tax to Financial Services.” Canadian Tax Journal 36 No. 5 (September– October, 1988): 1204–24. Hoffman, Lorey Arthur, Satya N. Poddar, and John Whalley. “Taxation of Banking Services Under a Consumption Type, Destination Basis VAT.” National Tax Journal 40 No. 4 (December, 1987): 547–54. Institute for Fiscal Studies. The Structure and Reform of Direct Taxation. Report of a Committee Chaired by Professor J. E. Meade. London: George Allen & Unwin, 1978. Meade, James, ed. The Stucture and Reform of Direct Taxation. London: Allen & Unwin, 1978. Poddar, Satya, and Morley English. “Taxation of Financial Services under a VAT: Issues and Options.” Presented at the National Tax Association Conference on The Taxation of Financial Services, Florida, 1994. Vanesse, Pierre. Sales Tax Reform and the Canadian Financial Services Industry: Exploring the Options. Financial Services Research Program Report for The Conference Board of Canada, Ottawa, March, 1989.

REFERENCES

Vanesse, Pierre. Operating as a Tax-Exempt Corporation: Canadian Financial Institutions and the GST. Financial Services Research Program Report for The Conference Board of Canada, Ottawa, November, 1989.

American Bar Association Section of Taxation. “1975 ABA Section Tax Report: The Special Committee on the Value-Added Tax.” Reprinted as “Technical Problems in Designing a 111