INVESTMENT INCOME TAX BASE
Income tax law around the world has been built up over the years in an ad hoc manner without clear design principles. The income tax law of many countries, when first drafted in the early twentieth century or before, focussed on the annual or recurrent receipts and costs of an investment. Not much attention was given to the changing value of associated assets perhaps apart from depreciation of some assets and the valuation of some trading stock. With an apple orchard, for example, included in the orchardists' income tax assessment would invariably be the proceeds of sales of apples from the orchard less the annual or recurrent costs of producing the apples, such as the cost of insect spray and pickers' wages. But the change in value of the orchard was often not taken into account in the measurement of taxable income.
Thus, these early attempts at taxing the income from investments were not really 'income tax' laws. They did not attempt to take into account a crucial component of income - the changing values of all the assets and liabilities associated with an investment. Usually the laws had no general definition of assets or liabilities. Over time, however, the handling of changing values of some business assets and liabilities have been grafted on to these fledgling income tax laws:
The end result of such grafting on of provisions dealing with changing asset/liability values is usually a complex and lengthy 'patch-work' of law with no overall guiding principles. Income tax laws are becoming lengthier and more complex despite clear movements towards change in value of assets/liabilities being included in tax assessments. Were a country looking to introduce its first set of business income tax laws, traditional law with its patch-work of 'add-ons' could not be regarded as a sound design template.
INCOME FROM ASSETS AND LIABILITIESWhat would be a suitable design template for laws to tax the income from investments? To answer this, first a clear understanding has to be established of what is annual income from business assets and liabilities. Given it is income from assets and liabilities that is to be subject to income taxation it would be a good idea to know what that is. A clear definition of income should be crucial to determining over-riding design principles of business income tax law. Having established these general principles, the drafting of the rest of the law would then take into account:
In all of this, defining the income from assets and liabilities is of central importance. Regardless of what collective entity might be holding assets/liabilities and regardless of where the income from the assets/liabilities flows, it is the assets/liabilities themselves that produce the income that is the focus of 'income' taxation.
That receipts from Jim the orchardist's sale of apples add to his income and the year-by-year costs of maintaining the orchard subtract from Jim's income is non-controversial. Similarly, if Jim bought the orchard for $1000 in 2009 and sold it for $1200 in the same year, people seem comfortable with the idea that the $200 of capital gain adds to Jim's income in that year. The $200 increase could result from both increased land value and the first crop of apples from the orchard being one year closer. Generally, however, people seem less comfortable with the proposition that even if Jim did not sell the orchard in 2009, Jim's income still increased by the $200 capital gain and could legitimately be included in his 2009 tax assessment. This lack of comfort remains despite the fact that Jim gets an income tax deduction for estimated annual depreciation on the tractor he uses during 2009 in the orchard even though he does not sell the tractor that year.
What arguments can be drawn on to help establish whether the 2009 increase in value of Jim's orchard is income even if he does not sell it?
First, all investors looking around for ways to make money know that capital gains are an important component of prospective returns from alternative investments. Prospective increases in share value have to be factored in with expected dividend flows. The rental property investor is interested in prospective increases in property value, not just the stream of rent payments. The issue is not whether capital gains are part of income but whether gains accrued, but not realised, in a particular year form part of that year's income.
The commercial finance sector is clear in its affirmation concerning the inclusion of accrued gains. Many financial firms include accrued gains and losses in their measures of income or profits across a wide range of financial products on a daily basis and make daily investment decisions in response. Others have more delayed responses but include accrued gains and losses in profit measures for decision-making purposes.
Secondly, a sound measure of income cannot depend on whether or not assets or liabilities happen to be given up in the year of income measurement. Income in the form of capital gains does not suddenly materialise just because disposal takes place - the accrued increase in asset value and therefore income has to be there before the increased sale value can be realised. This is not to deny that:
Nevertheless, the exclusion of accrued gains from tax assessment for reasons of practical political, measurement difficulties or value fluctuations should not be construed to imply that accrued gains do not form part of annual economic income.
Thirdly, sensitivities over accrued gains aside, of particular relevance is the observation that if accurate annual estimates of accrued changes in value of assets and liabilities are included in 'income' to be taxed, investors' pre-tax returns across alternative investments may be neatly reduced in proportion to investors' tax rates. Take an investor on a 47% tax rate - like Jim the orchardist - choosing between alternative marginal investments in, say, manufacturing, mining, agriculture or the financial sector all offering 10% per annum return before tax. After tax, this investor would be choosing between these same investments - like Jim's orchard - offering 5.3% per annum return (the pre-tax 10% reduced by the investor's 47% tax rate). Investors on a 30% tax rate would be looking across all these same alternatives offering 7% return after tax. This sets the scene for investors' decisions on where to put their money being pretty much the same after tax as before.
This remarkable outcome where taxing the income of investments has a neutral effect on investors' decisions is demonstrated mathematically (in net present value terms) by Mayo (1984) using discounting of investments' year-by-year cash flows consistent with year-by-year tax assessments. The outcome requires the not unrealistic assumption that, risk aside, the general 'going' pre-tax return of alternative investments (10% in the above pre- and post-tax comparison) is the interest rate used by investors to determine the year-by-year value of the investments (by discounting their future cash flows). Mayo (1984) extends the analysis with no change to the neutral investment outcome to incorporate risk. How this neutrality result comes about is made more tangible in the next section.
How good would it be for governments to be able to collect income tax to cover government expenditures without having a major effect on investment decisions? Consumption/savings decisions and work/leisure decisions may be affected but a neutral effect on investment decisions is offered. To achieve such an neutral outcome, however, not only would Jim the orchardist include in his annual income tax assessment annual depreciation on his tractor and his apple trees but any estimated annual increases in value of his orchard's land. Concerns and practical implementation issues associated with taxing accrued gains have to be weighed against the benefits of a neutral effect on decision making.
NEUTRAL TAXATION OF INCOME FROM INVESTMENTSTo re-cap, how would the income from investments be taxed in a way that left investors' decisions largely unaffected? Included in investors' annual income tax assessments would be the usual annual receipts (adding to income) and costs (subtracting from income) of the investments. In addition, there would be the estimated annual changes in value of the individual assets and liabilities associated with each investment. Increased values of assets add to income and decreased values subtract from income. Increased values of liabilities subtract from income and decreased values of liabilities add to income. In fact, because annual receipts and annual costs add and subtract during a year to investors' cash holdings, the treatment of these annual receipts/costs could be subsumed within the overall annual change in value of investors' cash assets.
Figure 1 summarises the income tax base offering tax collections with limited effects on investment decisions - the 'benchmark' tax base. The benchmark base includes the nominal annual change in value of assets and liabilities along with the usual recurrent receipts and costs. Nominal, rather than real, change in value is included on the assumption that - consistent with income tax systems around the world - nominal, and not real, interest income is included symmetrically in the income tax base (nominal non-private interest payments deductible and interest receipts assessable).
Figure 2 illustrates the neutrality property of the benchmark tax base. Pre-tax returns of 10% across alternative investments are reduced to 5.3% for taxpayers on a 47% tax rate and to 7% for taxpayers on a 30% tax rate.
Simply put, the benchmark treatment requires that: all income (including changes in values of assets/liabilities) be included in tax assessments in the year it arises.
The characteristics underpinning the benchmark income tax base lend themselves nicely to the important practical task of estimating annual change in value of assets and liabilities. Take an investment with an up-front (start of Year 1) value of $1000 arising from the future cash flows shown in Figure 3. Net receipts (annual receipts less annual costs) occur at the end of each year starting with $250 for Year 1 and declining by 15% in ensuing years. The sale price of the asset at the end of Year 5 is $444 reflecting the value at that time of the future declining net receipts stream. The up-front $1000 value matches the sum of the future cash flow from each of the 5 years of the investment discounted at 10% to the start of Year 1. In other words, someone paying $1000 up front (start of Year 1) for the investment and selling it for $444 at the end of Year 5 would earn a 10% per annum rate of return.
The cash flows in Figure 3 can be used to illustrate a method of estimating the annual change in value of assets (or liabilities) from associated cash flows. This methodology can also give tangible feel to the neutrality properties of the benchmark tax base.
Take the situation where an investor pays (at the start of Year 1) $1000 for an asset returning 10% per annum. If no net receipts occur during Year 1, the 10% annual return for that year has to be achieved by the value of the investment increasing by 10% to $1100 (see Figure 4). If, however, $250 of net receipts occur at the end of the year as in Figure 3, the $1100 value at the end of the year must decline by $250 to $850. Someone purchasing the asset at the end of Year 1 will not receive the $250 (the previous owner does). The asset's $850 value comes from the cash flows beyond Year 1. This method of estimating annual change in value illustrated in Figure 4 shows that, in net terms, the asset's value declines by $150 over the year, matching the 15% decline in net receipts underpinning values in Figure 3.
Another way of viewing the $150 decline in value is to note that for the value of the asset at end Year 1 plus the $250 of net receipts produced by the asset at that time to discount at 10% (the 'going' pre-tax interest rate available in the financial market) to the $1000 at the start of the year, the value at the end of the year must be $850. That is:
Figure 4 may be used to illustrate the mathematics underlying the specification of an income tax base with neutral effect on investment decisions. We have seen that with discounting at 10% before tax, the $1000 up-front value of the asset is determined by:
Before tax, the net present value (NPV) of the investment is zero with discounting at 10%. That is, value plus net receipts at year end discounted to start of year minus the $1000 investment cost equal zero. Therefore the rate of return of the investment is 10% (a 'marginal' asset yielding the same return the investor could get from the financial market).
Before tax, the investor could get 10% per annum interest income by putting his or her money in the financial market. Post-tax, the investor on a 47% tax rate gets 5.3% from the financial market after paying 47% on the 10% financial market interest. Thus, after income tax and risk aside, for an investor with a 47% tax rate, the discount rate for assessing investments reduces from the 10% before tax rate to 5.3%. In assessing the investment in Figure 4 after tax with annual change in asset value allowed for tax purposes (in this case, the $150 loss in value allowed as a deduction) and with the $250 of net receipts included in assessable income as usual, the Year 1 investment equation for this investor looks like this:
After tax, the net present value (NPV) of the investment is zero with discounting at 5.3%. That is, value plus after-tax net receipts at year end discounted to start of year minus the $1000 investment cost equals zero. Therefore the after-tax rate of return of the investment is 5.3% (the same post-tax return the investor could get from the financial market). The taxpayer's investment choices across different investment opportunities should not be much affected by such tax arrangements. Notably, this neutrality outcome does not require the asset to be sold at year end, just its change in value over the year to be included in the tax base. That conclusion applies equally to increasing values of assets as to decreasing values (and to changing values of liabilities).
The above mathematical analysis underpinning this neutrality outcome shows that the $1000 pre-tax value of the asset in Figure 4 stays the same after tax. The analysis shows that result occurs if tax reduces income, including change in asset value, by the same proportion - the investor's tax rate - that it reduces the pre-tax discount rate. Consistent with these requirements: (1) the benchmark (economic) income tax base to be subject to investors' tax rates incorporates annual change in asset value; and (2) income taxation, by including taxpayers' interest income, reduces investors' discount rates in proportion to their tax rates. As Swan (1976) notes on page 172, 'the neutrality of the tax can be seen to arise from the fact that while returns are taxed the opportunity cost of funds is reduced in the same proportion'. Sieper (1995, pg A1-ii) demonstrates mathematically the above asset value invariance over a single time period.
Invariant asset valuations with the taxing of economic income is shown in the general case (more than one period) by Samuelson (1964) with continuous discounting and by Mayo (1984) with discrete discounting corresponding to annual tax assessments. Mayo (1984) extends the analysis to include risk (essentially explaining how the spread of pre-tax possible outcomes - including investment failure - is 'squeezed' in a proportional fashion by a tax on economic income). Sieper (1995, pg A1-iv) provides mathematical proof, drawing on 'risk-adjusted' probabilities of different outcomes of an investment, of asset value invariance if income tax is based on realised economic income.
If market value of an asset (or, better, its value in its particular use) is readily available, the annual change in that value could feed directly into the benchmark base of Figure 1. If, however, such valuations are not readily available for a particular asset or liability but the future cash flows associated with it are, the approach in Figure 4 offers a practical way of estimating annual value.
In 'tax value' terms, the change in value estimated from the methodology in Figure 4 becomes ('tax value' is just a general term for the value for tax purposes, substituting for a variety of terms used for different assets, such as 'depreciated value' with assets attracting depreciation deductions and 'cost base' with assets subject to CGT):The asset depicted in Figure 4, acquired for $1000 at the start of Year 1 and declining in value by $150 over the year, is consistent with an item of plant or equipment whose net receipts stream is declining by 15% a year (perhaps because more and more maintenance expense is required over its years of use). In practice, even if an appropriate rates of return could be applied to items of plant and equipment, overall net receipts of a business could not usually be allocated accurately to each individual item of plant and equipment used by the business to produce those net receipts. Thus, change in value - or depreciation - for income tax purposes for plant and equipment is most often determined via a schedule of effective lives or write-off rates for different items of plant and equipment. Traditionally:
The general change in value methodology of equation (1) and Figure 4 works much better with financial assets and liabilities, including some leases and rights over other's assets or liabilities - where the net receipts or payments directly associated with the asset or liability are clearly specified and a suitable associated rate of return can be imputed (such as the internal rate of return of the financial flows involved). Figure 5 illustrates this for a zero-coupon discounted bond with a face value of $1000 payable at the end of Year 5 and purchased at the start of Year 1 for $621. At the time of purchase, the transaction offers a return of 10% ($621 compounded forward at 10% over 5 years equals $1000) and there are no annual net receipts - just the final payment of $1000. On the basis of equation (1), the tax value of the asset at the start of Year 1 equals $621 and its tax value increases (and that increase therefore included in the investor's tax assessment) at the end of each year by the prior tax value times the transaction's 10% return. At the end of the last year, the tax value equals the $1000 payment.
The structure of the cash flows of the discounted bond in Figure 5 is identical to any asset - such as an unimproved block of land with no associated net receipts - with an up-front payment and a final sale price. Either annual valuations of the land or application of equation (1) using a suitable percentage return could provide annual changes in tax values for income tax assessment purposes. Unlike the case of discounted zero-coupon bonds with their known final face value, however, governments are usually not comfortable with the idea of taxing accrued capital gains of assets like property or company shares.
This underlines the important point that the inclusion of accrued capital gains in the benchmark income of Figure 1 does not mean that such gains are generally included in income tax assessments. The shape of the tax law is the prerogative of governments. In the practical implementation of income tax laws, 'change in value' of assets/liabilities in Figure 1 becomes 'change in tax value' as governments decide how particular assets and liabilities are to be treated for income tax purposes. Rather than the annual change in value of assets/liabilities being included in tax assessments, the change in the government-imposed 'tax value' is. Thus, commonly the value for tax purposes of assets such as land and company shares is effectively left unchanged over the years that a particular investor own them - with the result that any tax on capital gains of such assets does not occur until the assets are sold.
Nevertheless, the 'pure' tax treatment summarised in Figure 1 (and demonstrated with Jim's orchard) with annual accrued change in value of assets/liabilities included in assessments represents a benchmark for the taxing of income from investments. It provides the conceptual basis for taxing accrued gains if, say, arguments over neutral tax outcomes or over equity in tax treatment so dictate. In fact, it provides the conceptual basis for taxing gains realised on the sale of appreciating assets if they have not already been taxed as they accrued, as well as for providing annual deductions for accrued capital losses on depreciating assets. Moreover, the benchmark base's proportional reduction in pre-tax returns illustrated in Figure 2 helps provide a basis for conjecturing on the impact on investment flows of taxing some assets/liabilities in ways that differ from the benchmark treatment. The tax treatment of plant and equipment with declining values over time might, for example, closely approximate the benchmark. In contrast, property and shares with commonly increasing values over time are usually not taxed on capital gains until sale - and then often not all the gains are taxed. Comparing computed after-tax returns of differently treated assets with common pre-tax returns can provide an understanding of the consequent pressures on resource flows (even though, in practice, price movements in response to these pressures - a symptom of non-neutral tax design - will tend to result in similar after-tax returns given similar risks).
If the early gains in the value of Jim's apple trees are not taxed, the post-tax return increases from 5.3% to 5.6% (ignoring price adjustments). This after-tax return increases much further to 8.4% (again ignoring price adjustments) if, as sometimes applies to selected agricultural pursuits, the up-front capital costs of establishing the orchard were immediately fully deductible. For illustrations of the distortive effects, including price adjustments, that can arise from concessional income tax treatment (accelerated depreciation and not taxing accrued gains year by year) see the discussion on negative gearing. For illustrations of the pressures that can arise from not even taxing the full realised gain when appreciating assets are sold see discussion on implications of entity taxation design of 'permanent' tax preferences.
THE 'TAX VALUE METHOD'A number of government-inspired reports in Australia reflect the growing recognition of the central importance to income taxation of changing values of assets and liabilities.
The Australian Government (1985) Draft White Paper, which was the precursor to Australia's general CGT system (and full imputation system of company taxation), looked in Chapter 19 at depreciation allowances that would more closely match the changing money value of durable assets under inflation and contrasted those allowances with depreciation allowances available at the time under accelerated 5/3 depreciation arrangements.
The Australian Government (1998) New Tax System, the precursor to the introduction of Australia's goods and services tax (GST), discusses business income taxation in Chapter 3. In that chapter, foreshadowed consultation on reform of the investment income tax base was to take into account a number of considerations, including the 'potential benefits of bringing tax value and commercial value closer together'. The chapter notes that 'closer alignment of tax value and commercial value is important with financial assets and liabilities in today's dynamic and innovative financial markets'. In the August 1998 terms of reference of the Review of Business Taxation headed by John Ralph, one of the considerations that the Australian Government required the Review to take into account was 'the potential benefits of bringing tax value and commercial value closer together'.
Some of the recommendations of the Review of Business Taxation (1999, Overview, Recommendations, Estimated impacts) relevant to the issue of tax value versus commercial value reflect a differentiation between, on one hand, traditional CGT assets such as real property and company shares and, on the other, general financial assets and liabilities (including various leases and rights over physical assets or financial assets/liabilities). For traditional CGT assets, the Review recommended (implemented by government) that taxation on a realisation basis remain and that the prior exclusion from assessment of an inflationary slice (initial cost base times annual inflation) be replaced by exclusion of half the capital gain - except for companies where no exclusion was recommended (full capital gain assessed but only on realisation). For financial assets and liabilities (including leases and rights), however, many recommendations involved practical ways of moving annual tax value towards annual commercial value. For example, the draft legislation of the Review - Review of Business Taxation (1999, Draft legislation, pg 145) - includes the following specification of the tax value at a particular time of a financial asset with known cash flows:
Moreover, the Review recommended a general 'cash flow/tax value approach' to determining taxable income (the so-called 'tax value method'). Reflecting that recommendation, the definition of net income for income tax purposes in the draft legislation of the Review (pg 57) is expressed as:
As noted, countries have over time commonly specified in their income tax laws the tax values of more and more types of assets and liabilities, such as: depreciating physical assets (with their changing 'depreciated values'); assets subject to CGT regimes (with 'cost bases' of those assets often remaining unchanged at initial asset cost or increasing with annual inflation rates until the assets are sold); and financial assets and liabilities (often with annual tax values changing on the basis of actual market values or values imputed from known cash flows of associated transactions). When previously excluded transactions are to be included, the method of determining tax value profiles of associated assets or liabilities would, as now, be specified on a case-by-case basis - hopefully with clear explanations of the reasons behind the method specified. Thus, the formulation of equation (4) is consistent with the practical design of existing income tax law.
Despite the presentational nature of equation (4), formally determining taxable income according to the formulation of equation (4) would have a number of advantages:
Thus, there are many practical reasons for basing the design of income tax law on the formulation in equation (4).
All of this would move beyond the practical benefits like simplicity of design and transparency and become more interesting if 'value' replaced 'tax value' in the equation (4) formulation in an all-inclusive system - or, in practice, if commercial value were the default under equation (4) for tax values of assets/liabilities without their own specific provisions in the law. This would immediately take the equation (4) - or Figure 1 - formulation from one offering simple consistency and transparency to one based on sound economic (and commercial) principle as well. The same drivers of commercial decision-making - change in value of assets and liabilities, as well as regular net receipts - would be underlined in the income tax base. Tax value methodologies for different groups of assets/liabilities would still be specified. There would, however, be an over-arching general recognition of the economic relevance of change in value in annual income measurement. In addition, with actual value the default for tax value there would be added motivation for clear explanations of tax value methologies for particular assets/liabilities that were not simply practical attempts at estimating annual change in value.
More likely to be achievable than such a commercial 'value' default would be an appreciation by policy advisers, policy makers and legislators of the advantages in terms of investment neutrality inherent in the 'benchmark' income tax base (Figure 1). Consistent with what is already happening with patch-work tax systems, such appreciation would mean that provisions determining annual change in tax value for particular asset/liability types would vary largely according to: (1) the most practical ways of estimating annual change in value; and (2) the degree of political acceptability of attempting to include annual change in value in tax assessments for particular types of assets/liabilities. Appreciating assets, like land and company shares, currently subject to CGT systems, might not be any more likely to have all their capital gains included in income tax assessments when the gains are realised - or to have their accrued gains included in annual assessments. But there would be general appreciation that even taxing the full gain on realisation was concessional compared to the neutrality benchmark. Moreover, as financial engineering such as securitisation of tangible assets further blurs the distinction between physical and financial assets, commercial reality might be expected to enable laws to relax more towards the benchmark.
Version 1.1 © Copyright Wayne Mayo 2009