The Macaulay Duration : A Key Indicator for the Risk-Adjustment in Fair Value

International Financial Reporting Standards (IFRS) 13 Fair Value Measurement lays down two methods to adjust Expected Present Value (EPV) for risk. According to Method 1, expected cash inflows should be risk-adjusted by subtracting a risk-premium and discounted at the market risk-free rate, see (IFRS 13, B25). In contrast according to Method 2, expected cash inflows should be discounted at the risk-free rate augmented by a risk-premium addendum, see (IFRS 13, B26). Standard IFRS 13, B29 leaves the freedom to choose between the two methods. The aim of this note is to identify the relationship between the Risk-Adjusted EPVs rolled out from Method 1 and Method 2. First we introduce a theoretical solution to risk-adjustments compliant with the Standard IFRS 13, B29. Then, we set up a user-oriented proxy to connect the risk-premium present in Method 1 with the risk-adjusted rate present in Method 2. This proxy spots light on the key role played by the Macaulay Duration of expected inflows, rather than that of the lifetime of the project. As a consequence, projects expiring at the same redemption date and endowed with the same EPV and/or the same total inflow may differ considerably in risk-adjustments, due to different Macaulay Durations. A user-oriented method to properly to fast evaluate risk-adjustments for multi-cash inflow projects is provided. Sensitivity analysis of the impact of the Macaulay Duration on Risk-Adjusted EPV is also rolled out through numerical examples.


Introduction
International Financial Reporting Standard 13 (IFRS 13) Fair Value Measurement defines the fair value as the 'price that would be received to sell an asset or paid to transfer a liability in an orderly transaction in the principle (or the most advantageous) market at the measurement date'.Since January 2013 IFRS 13 is mandatory.Fair Value View is the most effective way to satisfy lenders and investors needs of information in their process of financial resources allocation, as stated in Whittington (2008Whittington ( , 2015)).Their needs will be served because: 'Investors', lenders' and other creditors' expectations about returns depend on their assessment of the amount, timing and uncertainty of (the prospects for) future net cash inflows to the entity.' (Conceptual Framework OB3).
IFRS 13 defines fair value as "the standard" in market valuation and examines consequences from the application of this "standard", when, in valuating assets and liabilities, no directly observable market information is available.
This paper analyzes how it is possible to create a correspondence between different methods in order to determine Expected Present Value (EPV) under conditions of uncertainty, within the fair value (FV) definition.
Since in the Conceptual Framework for Financial Reporting (2010Reporting ( , revised in 2018)), jointly designed by IASB and IFRICs, comparability is established as one of the qualitative characteristics of useful financial information, this article applies the principle within the methods for the definition of the EPV when risk adjustment techniques are required.
IFRS 13 establishes that, without directly observable market prices, the need is to proceed through valuation models.However, those financial models must take into account all inputs directly observable from the market, in accordance with the "fair value hierarchy", which results in a market-based, rather than entity-specific, measurement.As part of this assumption, in IFRS 13 a differentiation is envisaged in relation to the evaluation method that can be used to determine the FV (Paragraph 62).
Therefore, it is crucial to recognize a non-objective dimension in the principle, reflected in the choice the individual will take in selecting a model to determine the FV, out of the market evaluation.It addresses the issue of how to include the risk profile in determining the instrument financial value.Specifically, in the "income approach" for the EPV determination, the evaluator deals with uncertainty issues that in the market are assessed by the risk premium.
The provision of Method 1 and Method 2 (B25 and B26) for measuring the risk premium on expected cash flows displays a significant issue, leading to the review of the logical architecture of the Conceptual Framework, if any kind of comparison should not be feasible.
Besides comparability, some other qualitative characteristics for useful financial information are relevant.KPMG (2010) highlights that: "The Framework provides a broad discussion of the basis of preparing financial statements.It discusses the objectives of financial statements; their underlying assumptions and qualitative characteristics, such as relevance and reliability."And it adds that: "In many cases there is a trade-off between relevance and reliability of information…".
We specifically face this trade-off in the determination of the EPV.In fact, according to Method 1, the determination of the EPV proceeds by adjusting expected cash flows for the risk premium and then realizes the discounting process by applying a risk-free rate.According to Method 2, the EPV is determined by discounting unadjusted risky cash flows by means of discount rates that incorporate the risk premium required by market participants.
The existence of the two Methods and the required reliability for evaluations lead to a comparison of the assumptions on risk adjustment techniques (B17).IFRS 13 establishes that must be evident the link existing between the cash flows and the discount rate.Thus, if the considered cash flows are generated on a contract basis, similar to ones of other deals in an efficient market, it is correct discounting with the implicit rates of the contracts already traded on the market.If the cash flows are produced by an activity that does not have similar financial profile with other market traded contracts, a need to estimate a specific risk premium for the activity arises, in order to maximize its comparability with other activities traded on the market.This issue arises when the cash flows are not fixed in the contract, but result from the operators' expectations.It set a realistic problem of comparability among institutions' financial statements.
For all these reasons, it is crucial to have a comparison tool, which allows to attain the compliance of Methods 1 and Method 2 with IFRS 13.
The contribution of this article supports this aim.The Macaulay Duration is just that tool to compare the results of Method 1 and Method 2, understanding and deepening the reasons for any differences.
First we state a general Risk-Adjusted EPV formula complaint with IFRS 13, B29.
Second, we provide a proxy solution that spotlights the direct relation between the cash risk-premium used in EPV Method 1 (IFRS 13, B25) and the risk-premium rate used in EPV Method 2 (IFRS 13, B26).
We show that the Macaulay Duration of expected cash inflows, rather than the lifetime of the project, is the key driver in risk adjustments.
The remainder of the paper is organized as follows.In Section 2, we discuss the Standards IFRS 13, B25 and IFRS 13, B26.In Section 3 the compliance condition with IFRS 13, B29 is settled.In Section 4 main results are discussed.In Section 5 the results are illustrated using numerical examples.Section 6 concludes the note.

Method 1 and Method 2 for computing Risk-Adjusted EPV
IFRS 13, B17 commands to adjust EPV for risk according to one of the two risk-adjusting techniques: • EPV Method 1 (IFRS 13, B25) adjusts the expected inflows by subtracting a cash risk-premium.That method is grounded on the actuarial certainty-equivalent principle that commands that uncertain expected value should be depreciated by a risk-premium, whose amount depends on the decision maker risk aversion.Finally, risk-adjusted cash inflows are to be discounted at the market risk-free rate.
• EPV Method 2 (IFRS 13, B26) prescribes to discount the expected cash flow at the risk-free rate augmented by a risk-premium addendum.
Standard IFRS 13, B29 leaves free to choose between Method 1 or Method 2 to adjust EPV for risk and compute the so called fair value.
We lay out the notation used throughout the paper.Let • 0 s a ≥ is the expected value of the uncertain cash inflow of the project A, at time s t , 1,..., s n = ; • i is the market risk-free rate (Note 1); • r is the risk-premium rate; • ( ) • ( )  is the NPV of the risk-premium s p cash flow discounted at the market risk-free interest rate i; is the Macaulay Duration discounted at the market risk-free rate i (see Macaulay, 1938; for a discussion on the use of bond duration see for example Fabozzi, 1999) (Note 2).For simplicity but without loss of generality, we assume initial time 0 0 t = .

EPV Method 1 (IFRS 13, B25)
EPV Method 1 (IFRS 13, B25) relies on the framework of the Certainty Equivalent, widely discussed in the literature (see Fishburn, 1986;Chen, 1967;Mao, 1970).The expected inflow s a is diminished by the cash risk-premium s p , for 1,..., s n = . Under the assumption that the uncertain inflows of the project A are stochastically independent (Note 3), the Risk-Adjusted EPV can be simply calculated as the sum of the discounted risk-adjusted expected cash inflow ( ) According to EPV Method 2 (IFRS 13, B26), the expected cash inflows s a , 1,..., s n = should be discounted at the risk-free rate i augmented by a risk-premium addendum r.It results where i+r is the risk-adjusted rate of interest.
Method 1 and 2 lead to the same Risk-Adjusted EPV value if (and only if)

( ) ( ) ( )
(1) Equation ( 1) can be called the condition for compliance with IFRS 13, B29.In general, formula (1) cannot be solved to come up an explicit expression of the risk-premium rate r as a function of the cash-risk premium s p , for 1,..., s n = . A user-oriented proxy of such relation will be achieved in Section 3.

The role of the Macaulay Duration in Risk-Adjustments
First, we tackle the case of projects with a single-cash inflow.The both explicit and a proxy formula for the Risk-Adjusted EPV will be achieved.

Projects with a single-Cash Inflow
Let the project has a single-cash inflow 1 a cashable at maturity 1 t .For 1 n = the compliance condition (1) comes down to ( ) ( ) The risk-premium becomes Discounting both hand-sides of above at the market risk-free rate i, we obtain ( ) ( ) ( ) So, we achieve the exact expression for ( ) The risk-adjusting factor ( ) holds the following properties: • 1 ϕ < .That informs that the Risk-Adjusted EPV is ϕ per cent of EPV.For example, if 0.80 ϕ = that means that the Risk-Adjusted EPV is 80% of EPV; • ϕ is an increasing and convex function of the risk-free rate i.Therefore, the higher the risk-free rate i, the higher ϕ and the lower the impact of risk-adjustments.Due to the convexity of ϕ , the marginal effect is increasing; • ϕ is a decreasing exponential function of the cashable date 1 t .Therefore, the longer time the cashable date 1 t , the lower ϕ and the stronger the impact on risk-adjustments.
The following linear approximation of (3) shows up the key drivers in risk-adjustments see the proof in the Appendix A. If there is a single expected cash-inflow 1 a cashable at 1 t , the Macaulay Duration 1 D t = .Above can be rewritten as In the special case of 1 1 D t = = year, the risk-premium becomes and the exact expression (3) becomes And the linear approximation

Projects with Multi-Cash Inflows
Although the condition for compliance with IFRS 13, B29 can be always solved numerically, a general solution for multi-cash inflow projects is not available in closed form.
Analogously to the single-cash inflow projects, we give a linear approximation of the Risk-Adjusted according to Method 2, ( ) ( ) Where ( ) discounted at the market risk-free rate i (see the proof in the Appendix B).Note that D is a decreasing and convex function of the risk-free rate i. so as the risk-free rate i increases, the risk-adjustments decrease.

Main Results
We are now ready to sum up our main results.

Result 1
NPV of risk-premium cash flow ( ) P NPV i used in Method 1 (IFRS 13, B25) and risk-adjusted rate r used in Method 2 (IFRS 13, B26) are approximately related as follows ( ) where ( ) is the Macaulay Duration discounted at the market risk-free rate i.
Remarks on formula (8) follow: a.One might think that projects with the same redemption date and equal EPV or equal total inflow should be risk-adjusted by similar cash-premium amounts.That conjecture is not true in general.In fact, risk-adjustments are not proportional to the lifetime of the project, but rather to the Macaulay Duration D of expected inflows (see Examples 1 and 2 in Section 5).
b. Projects with the same EPV and the same Duration D are risk-adjusted by approximately the same cash risk-premium NPV.
c.The relative ( ) Where D plays the role of cash risk-premium elasticity in emphasizing or smoothening the EPV risk-adjustments.

Result 2
Risk-Adjusted EPV complaint with IFRS 13, B29 can be approximated by where ( ) ( ) is the Macaulay Duration (Note 4) discounted at the market risk-free rate i.Formula (10) makes evidence that the Macaulay Duration ( ) = summarises all information needed for risk-adjustments.

Result 3
Risk-Adjusted EPV approximation ( 10) is always rounded down.The approximation is rather good when the risk-premium rate r is small, on the contrary in the presence of large risk-premium rate r, the quality of the approximation deteriorates.The error magnitude is amplified or smoothened by D. So, cash risk-premium approximation is good for projects with short Duration, but it may flaw down for projects with long duration.Goodness-of-fit of will be tested with simulations in Section 5, Example 4.

Numerical Illustrative Examples
To ascertain the relevance of the Macaulay Duration on risk-adjustments we set a number of illustrative examples.Throughout this Section we assume that the market free-risk rate is i=5%.

Example 1: Projects with equal total inflow and different Macaulay Duration
Let the projects A, B, C, D be structured as in Table 1.Projects A, B, C, D expire in 4 years, have total inflow of €2,000 but display different D.
The most significant inflows (i.e.those of expected amount €1,000 and €800) are: • at the beginning of the lifetime period for the project A; • at the extreme dates of the lifetime period for the project B; • about at the middle of the lifetime period for the projects C and D. The Macaulay Duration of the projects A, B, C and D goes from 1.6201 years to 3.3186 years.For short D (as for D=1.6201 years) the relative impact of the NPV of cash risk-premium on EPV is contained (of 4.56%), in contrast for high D (as for D=3.3186 years), the relative impact of the NPV of cash risk-premium on EPV is strong (of 9.77%).In conclusion, even Projects A, B, C and D have the same total inflows and maturity, risk-adjustments are very different, because of different Macaulay Durations.

Example 2: Multi-cash inflow projects with equal EPV and different Macaulay Durations
Let the projects E, F, G and H be structured as in Table 2. Projects E, F, G and H expire in 4 years, have EPV of about €1000 and have the Macaulay Duration of about 1, 2, 3 and 4.
2 In conclusion, even Projects E, F, G, and H have the same EPV and maturity, risk-adjustments required by IFRS 13, B29 differ considerably.The impact of the NPV of the cash risk-premium on EPV goes from 2.86% to 11.92%.

Example 3: Single-cash inflows with and different Macaulay Durations
Let now consider single-cash inflow projects I, L, M and N with EPV of €1000, collected in Table 3.These projects have the same EPV and Macaulay Duration of the multi-cash inflow projects E, F, G and H, collected in Table 2.
Data in Table 3 make evidence how the single-cash inflow projects I, L, M and N replicate the multi-cash inflow project E, F, G and H, respectively, that has the same EPV of €1,000 and the same Duration (Note 5).
That shows that any multi-cash inflow project can be replicated by a proper single-cash inflow project.Risk-Adjusted EPV approximation ( 10) is always rounded down respect to the exact solution of equation ( 1), because NPV is a decreasing and convex function of the interest rate used in discounting.Goodness-of-fit is sensitive to the scale of the risk-premium rate r.Data in Table 4 show that approximation is rather good when the risk-premium rate r is small (around 1%), while it may deteriorate for higher r.Approximation error is also sensitive to D. For long D the absolute error is amplified, for short D the absolute error is smoothened.
To improve the goodness-of-fit is sufficient to go on with higher order approximations.The second order approximation involves the so-called (bond) convexity index (for the notion of bond convexity, see for example Fabozzi, 1999).

Conclusions
Standard IFRS 13, B29 leaves freedom to risk-adjust EPV with Method 1 (IFRS 13, B25) or Method 2 (IFRS 13, B26).A general formula that makes these two methods equivalent, is set up.Although this formula is always numerically computable, a closed-end expression that relates cash risk-premium present in Method 1 with risk-premium rate present in Method 2, is not achievable for multi-period projects.So we settle a user-oriented proxy that spotlights on the key drivers in risk-adjustment methods.
The main result of this note is that we make evidence that the Risk-Adjusted EPV is approximatively proportional to the Macaulay Duration of the expected inflows, rather than to the lifetime of the project.It follows that projects expiring at the same redemption date and having the same EPV and/or the same total inflow, may considerably differ in risk-adjustments.must be expressed in the same unit of time, i.e. the commercial year composed of 360 days, or the effective year composed of 365 or 366 days.
Note 2. Note that the Macaulay Duration is an approximation of the implied duration recently introduced by Bornholt (2017) in the field of bipole projects.Note 3.Under the assumption of stochastic independence among uncertain cash inflows, upper and lower bound for EPV can be found using the notion of copula (see McNeil et al, 2005).Note 4. Formula (10) informs that risk-adjustments are approximately proportional to the yearly risk-premium rate r and the time-weighted average D expressed in years.The financial factor ( )  plays the role of the linear bank discount factor used in Financial Mathematics to calculate the premium for discounting a bank bill, where D is the maturity of the equivalent single-inflow project.Note 5. Note that the Macaulay Duration of single-cash flow projects coincides with the cashable date of the single-inflow.
discounted at the interest rate i.It is also called the Expected Present Value (EPV) of the project A; -

Table 1 .
Projects with total inflow of €2000 and different Macaulay Duration D

Table 2 .
Projects with similar EPV and different Duration D

Table 3 .
Risk-adjusted EPV of single-cash projects with EPV of €1,000

NPV of cash risk-premium % on EPV 2.86% 5.80% 8.82% 11.93% Example 4 Goodness-of-fit of the Risk-Adjusted EPV approximation
To test the significance of the approximation (10) we calculate the approximation error at different risk-premium rates for the projects A, B, C, D defined in Table1.

Table 4 .
with risk-premium rates , and .
Illustrative numerical examples support our findings.