Accounting Control

By Humphrey Nash

One defining characteristic of an asset is control.  If one does not have control of an asset then the asset represents a trust or a hope, not a current reality.  Lack of control often means events or forces may interfere with the realization of the asset.  Lack of control implies uncertain outcome and potential impairment of the asset.  It seems that control of an asset is important prerequisite to accounting recognition.

But is it?

We may wish to have control over an asset; we may pretend to have control; we may even expect to have control.  In fact, our control is limited.  The safest government bond, for example, only gives the illusion of control.  The market value of a bond can fall with rising interest or inflation rates, rates controlled or uncontrolled by government actions.  The bond’s maturity value in real dollar terms may be a fraction of its real purchase price.

Assets are a claim on an unknown and variable future; the future only has current value as an expectation.  We must somehow control both the future and our expectations of the future, at least enough to satisfy the definition of an asset.

by a particular entity as a result of past transactions or events.

This definition recognizes both the need for control and our limited ability to control.  The degree of control need not be absolute, merely enough to make realization of the asset "probable".  This is a crucial distinction because it permits accounting to value an inevitably uncertain future.  This permits accounting to look forward and to support decisions.  FASB has made great progress in exposing accounting to the ideas of a variable future with a range of possible outcomes and the use of probability and interest discounts as a means of measuring the value present in expected cash flows.  How does control coexist with an uncertain future?

It is instructive to look at a few simple examples of potential assets with varying degrees of control and then to ask what level of control is appropriate in defining assets.  It will also be instructive to consider who should define "control" and what level of control is appropriate for them.

Example 1.  I purchase a lottery ticket at a cost of $1 today.  Tomorrow’s cash jackpot is $10,000,000.  The value of the ticket tomorrow will be 0 or $10,000,000.  The odds of winning are 1 in 20,000,000. What is the today-value of the lottery ticket?  The (pretax) expected value of the two possible outcomes is,

.00000005 * $20,000,000.00 +.99999995 * $0.00 = $0.50

This value is reasonable and the accountant might be willing to assign an asset value of $0.50 to the lottery ticket.  We have an acceptable value despite an extreme lack of control.

Example 2.  I can today-purchase the right to $1,000,000 based on the tomorrow-flip of a coin.  What is the today-value of the flip right?  One answer that pops into mind is $500,000, the expected value.  How much would I pay for the flip right?  I doubt that I would pay more than $100,000.  The me-value of the flip right is $100,000.  For me, losing a sure $100,000 is worse than losing an expected $500,000 or a possible $1,000,000.

Example 3. A large corporation can today-purchase the right to $1,000,000 based on a tomorrow-flip of a coin.  What is the today-value of the flip right?  One answer that pops into mind is $500,000, the expected value.  How much would the corporation actually pay for the flip right?  I expect that the corporation would pay more than $350,000.  The corporation-today-value of the flip right is $350,000+.  For the corporation with great risk capacity and tolerance, losing an expected $500,000 or a possible $1,000,000 is worse than losing a sure $350,000.  If the corporation purchases the flip right, what should its accounting value be?  I suspect that most accountants would be uncomfortable with a $500,000.  They would be almost as uncomfortable with a value of zero, especially if the actual cost was $350,000.   Should the accountant or auditor use his own comfort level as a guide, or the company’s, or any comfort level?

Example 4. A large corporation can today-purchase the right to $10,000 per flip based on the 100 tomorrow-flips of a coin.  What is the corporate-today-value of the flip right?  One answer that pops into mind is $500,000 (=100*.5*10,000), the expected value.  How much would or should the corporation pay for the flip right?  I expect that the informed corporation would pay more than $450,000.  The corporation-today-value of the flip right is $450,000+.  For the corporation with great risk capacity and tolerance, losing an expected $500,000 or a possible $1,000,000 is worse than losing a sure $450,000, especially since the probability of an outcome of less that $400,000 is now very low (.0176 to be exact).  The informed accountant would be quite happy to assign a value of $400,000 since the probability is 98.24% that the actual outcome will equal or exceed that amount.  This is about as certain as financial outcomes get.

What is the difference between Examples 3 and 4?   Both situations depend on coin flips, have an expected value of $500,000, with uncertain outcomes. The expected outcome variance is much larger in Example 3.  In Example 4 the outcome is essentially confined to the $400,000 to $600,000 range.  The difference results from the number of trails.

Through repeated trials corporations gain outcome control.  Most corporations sell a variety of products or services to a large number of customers so that Example 4 is a closer characterization of the level of outcome control that corporation typically achieve.  This has interesting implications for the technology of prospective accounting.  FASB has exposed the theory of Present Values of Expected Cash Flows (PVECF) in an accounting setting, but has maintained that control must also be present.  At the micro level, a single coin flip exhibits extreme lack of control, but, at the corporate or macro level, even coin flips can be adequately controlled.  This means that PVECF can, in practice, be an appropriate basis for asset (or liability) values because essential outcome control is normally achieved at the corporate scale.

In fact, PVECF exhibits other dimensions of control.  Since cash flows are recognized over several years the number of trials is increased creating more stability.  The PVECF time frame also allows management to learn and adapt, so that bad outcomes can be minimized or avoided.  PVECF may also involve a great variety of trials (diversification) so that uncertainty* is reduced and control is increased.

Within Accounting For The Future (AFTF), the PVECF technology is strongly disciplined (controlled).  AFTF depends upon and requires explicitly validated expected cash flow models.  Expectations are disciplined by the AFTF dual validation, which automatically discounts unduly optimistic management projections.

There is no better way to represent and understand the future that to create models of that future.  There is no better way to control that future than to understand it.  AFTF requires expected cash flow models for both accounting and decisions.  The financial decision process and the accounting valuation process are unified so that both processes support and discipline each other.

PVECF and prospective accounting posses several real dimensions of control.  Enough to satisfy a reasonable accounting requirement.

Financial reporting to shareholders (and their representatives: management, portfolio managers and analysts) should be made relevant to shareholders.  Management must take into account the general expectations and risk tolerance of the shareholder.  The scale, relevance, and utility of financial reporting must be judged from the shareholder’s perspective.

Example 5. An investor owns 1% of each of 100 companies.  Each company has a single asset: a $1,000,000 single flip right.  Each of the 100 companies provides an expected gain to the investor of $5,000 so that the total expected gain for the investor is $500,000.  As in Example 4 the actual outcome is essentially confined to the $400,000 to $600,000 range.  The investor’s diversification reduces his risk and increases his risk tolerance.  It is this risk tolerance which financial reporting should address.

Hence even a case of extreme lack of corporate control, such as a single coin flip, is acceptable to the diversified investor.  What the diversified investor wants to see is expected values since his diversification limits his outcome risk.

The coin flip example may seem unrealistic but more realistic cases may exhibit even more convergence to a stable and controlled outcome.

Example 6 (Insurance). The annual probability of death is .001.  The gross annual premium charged for a $1,000,000 insurance policy is $2,000.  The expected annual gain is $1,000 (= $2,000 - .001*$1,000,000).  The actual outcome for a single policy after one year is either a gain of $2,000 or a loss of $998,000.  An accountant would be naturally reluctant to attach a value of $1,000 to a single policy issue.  In the normal scale of insurance operations an insurer will have tens of thousands or hundreds of thousands of policies in force, but even with as few as 1,000 similar policies the outcome is very reliable, as shown in the table below.  There is less than a 10% chance of any loss at all and the probability of losing more than $1,000,000 is less than 1/2 of 1%.

The convergence to the expected outcomes in the above example is rapid.  In practice, convergence to expected outcomes provide a level of control that obviates the need for any risk disclosure in financial reports, even for the un-diversified investor.

Would the investor be better off if the probabilities or expected gains were understated?  No. As a rule, it is financially better to be fully informed.  Conservatism conceals information and produces sub-optimal results.  Suppose, for example, the conservative accountant assessed the flip right asset at $250,000 rather than the expected gain of $500,000.  This seems good or at least not harmful to the shareholder.  However, this harms the potential shareholder, who will bypass perhaps the best investment opportunity, and the current shareholder who divests.  The company can be harmed in that its equity cost of capital increases.  Conservatism impairs capital market efficiency.

Risk Aversion

Risk aversion is a relative concept.  Some investors are more risk-adverse than others.  By definition half are less risk-adverse than average.  Likewise half of publicly-traded companies are less risky than average.

Some investors may seek variance and  pay a premium for it.  The person who buys a lottery ticket has a poor expected return; he buys the ticket for its unexpected return.  He revels in outcome variance: he would not buy the lottery ticket if the payoff were a certain 50 cents.  Likewise, for two companies with equal expected gains as in Examples 3 and 4, an investor may prefer to invest in the more risky, i.e., to not be constrained to a return between $400,000 and $600,000.  Risk preference may be common in the equity markets.  The risk averse investor may prefer bonds.

If risk aversion dominates the capital markets then the diversified investor or the diversified mutual fund can achieve superior results by seeking out risk and taking advantage of the risk aversion and lack of diversification of others.  In an up-market a high beta (risky) stock will outperform the low beta stock; since the market tends to rise, risk may be preferred.  Ultimately, overall market risk aversion is self-extinguishing.

How should accounting cope with the variety of shareholder risk preferences/aversions, risk capacities, hedging positions, and portfolio diversifications?  One way is to remain completely neutral and base values on expectations with no bias against variance or toward conservatism.

Disclosure

There may be a better way, suggested by the SEC’s long standing position that risk assumption is a choice and responsibility of the investor and that the only duty of financial reporting is to fully disclose that risk.  Hence, the SEC (unlike some states) permits registration of risky or even speculative securities, so long as the risk or speculative nature is disclosed.  In Examples 3 and 4 the expected values are the same; the variance measures the degree of outcome control.  From a statistical standpoint outcome distributions are essentially characterized by the mean(expected value) and standard deviation(outcome variance).  From a financial reporting perspective, shareholder value based on discounted expected values together with disclosure of the outcome risk provides essentially all the information needed for investment decisions.

The AFTF solution

In the draft proposal Accounting For The Future, outcome variance disclosure was provided for in the sample financial reports (see Appendix 3:Financial Reports). The exhibit of Sensitivity of Company Valuation Due to Adverse Deviation shows the effects that deviations in cash flow components have on the reported value of the company.  These effects are the changes in the company value emerging from the cash flow model when a standard deviation change of each major cash flow is input one at a time.  The input standard deviations might be based on the historical cash flow record.  One possibility is to calculate the standard deviations over the 5-year validation period as modeled less actual.  This would be an incentive to develop a close fitting model and would limit calculations and model development to a recent five-year historic period.  If management can’t reasonably model the past, this would show up in the sensitivity analysis, as a more variable future  ...  as it should.

An un-diversified investor would do well to consider the sensitivity analysis.  The diversified investor or mutual fund doesn’t need to be so cautious.

There is more.

The capital markets assess the corporation’s risks and their own risk tolerance and charge accordingly.  Within AFTF the discount rate is the cost of capital which includes those risk charges.  It may not be necessary to further discount for risk or even to disclose risk since it is already factored into valuations.  I don’t recommend no disclosure since management may have special or new knowledge of factors and this knowledge may not yet have been incorporated in prices and the cost of capital.  The AFTF model will adjust values to the capital market’s risk assessment and risk tolerance.  If the capital market is risk adverse it will appropriately discount companies with more variable results.

Risk aversion may vary from company to company, from industry to industry, and may wax and wane with financial or economic conditions.  A universal risk premium, such as the CAPM model suggests, is unresponsive.  The capital market’s risk premium must depend on the capital market’s current assessment of risk as revealed by price.  AFTF uses capital market pricing to determine the appropriate discount rate.  If outcome control is important to the capital markets, AFTF will get the message and deliver the message.

A General Comment

From a more general societal or economic standpoint, the pursuit of certainty is a dead end.  Society must encourage diversity and variability so that evolutionary forces may act.  A centralized or planned economy creates certainty but stifles progress.  The current cost-based accounting model worships the past.  This guarantees certainty but discourages change and innovation.  Accounting should be forward-looking and should be, at least, neutral in regard to outcome uncertainty (variance).  To idolize certainty and require absolute control is to avoid the future.  This is not decision useful.

Conclusion

Is control of an asset an important prerequisite for accounting recognition?  Perhaps.

However, we usually measure and report at the corporate scale where many assets, which may be individually uncontrolled, are collectively well controlled.  The corporate level of control is sufficient to provide decision useful measures, for management, shareholders, and for an accounting model.  This level of control supports reasonable expectation and satisfies the definition of an asset.

*There are two types of outcome risk: 1. Variance from expected due to limited trials when the underlying probabilities are known and 2. Variance from expected because the underlying probabilities are unknown. Corporations manage the first outcome risk through repetition and manage the second risk through experience, research, and diversification.