Probability and Timing in the Valuation of Claims
A claim's headline value does not, by itself, tell us what the claim is worth. Before any recovery is realised, the matter may have to pass through jurisdiction, liability, quantum, challenge and enforcement. Even where each stage appears more likely than not, the probability of completing the entire sequence can be much lower than intuition suggests.
Timing adds a second complication. A recovery expected years from now must be discounted, but an award carrying interest continues to grow while payment is delayed. Whether additional time increases or reduces present value depends on the relationship between those two rates. The boundary between them is not a matter of instinct: it can be calculated.
Every stage probability is conditional on the one before it
Consider an illustrative claim that must clear five stages before cash arrives: the tribunal must accept jurisdiction; liability must be established; quantum must land at or above the modelled level; the award must survive challenge; and it must then be enforced in full. Assign each stage an illustrative probability of 0.85, 0.70, 0.60, 0.80 and 0.75 respectively. Illustrative parameters. Not calibrated to any actual claim.
Not one of those is a pessimistic assumption. The weakest is a coin-flip with the odds tilted in the claimant's favour; three of the five are 0.75 or better. The probability of reaching the end of the sequence is their product, which is 0.214 — a little over one chance in five.
The reason intuition fails here is that people read a sequence by its typical member rather than by its product. Presented with five numbers averaging 0.74, the mind returns something in the neighbourhood of 0.74. The arithmetic returns 0.21. Our published note on stochastic modelling makes the general point about compounding; what matters for valuation is the discipline that follows from it.
That discipline is conditionality. The quantity to estimate is not the probability of a particular quantum outcome considered in isolation. It is the probability of that outcome once liability has been established — and the distinction is not pedantic, because the two estimates can differ materially. A matter in which liability has been established is not a random draw from the population of matters; the facts and findings that carried it past liability are the same facts and findings a tribunal will weigh on quantum. An analyst who takes a merits opinion on liability, an expert's unconditioned quantum range and a separate enforcement view, and multiplies them together, has counted the same good news more than once. The resulting estimate is too high, but the source of the overstatement may not be immediately apparent.
The discount rate can materially alter present value
Suppose the sequence above resolves and produces an expected nominal recovery of ₹100 crore, five years from today. What is that worth now?
| Discount rate | Present value of ₹100 crore in five years |
|---|---|
| 5% | ₹78.4 crore |
| 8% | ₹68.1 crore |
| 10% | ₹62.1 crore |
| 12% | ₹56.7 crore |
Illustrative parameters. Not calibrated to any actual claim.
Same claim, same date and same nominal figure — but a difference of more than ₹21 crore between the highest and lowest present values in the table. The discount rate is not a rounding adjustment applied at the end of the analysis. Over a five-year horizon, it can be one of the most significant determinants of the valuation and should be stated alongside the result it produces.
Award interest can offset the erosion of present value
Elapsed time does not only push value down. Where an award carries interest, it also pushes the nominal amount up.
Under the Arbitration and Conciliation Act, 1996, section 31(7)(a), a tribunal may include pre-award interest "at such rate as it deems reasonable" unless the parties have agreed otherwise. Section 31(7)(b) goes further for the period after the award: "A sum directed to be paid by an arbitral award shall, unless the award otherwise directs, carry interest at the rate of two per cent. higher than the current rate of interest prevalent on the date of award, from the date of award to the date of payment." The expression "current rate of interest" takes its meaning from section 2(b) of the Interest Act, 1978.
So over any period of post-award delay, two things happen at once. The nominal sum grows at the award rate. Its present value shrinks at the discount rate applied. Which effect dominates is a question of arithmetic rather than judgement — and the arithmetic has a clean answer.
The point at which more time stops adding value
Take a claim with face value V that accrues simple interest at an award rate ra. A party valuing the award applies a discount rate rd, with annual compounding. After T years, the nominal amount is:
V (1 + raT)
Its discounted present value is:
PV(T) = V (1 + raT) / (1 + rd)T
To find the point at which present value stops rising and begins to fall, differentiate the logarithm of PV(T) with respect to T:
(d / dT) ln PV(T) = ra / (1 + raT) − ln(1 + rd)
Setting this equal to zero and solving for T gives:
T* = 1 / ln(1 + rd) − 1 / ra
Here, T is the number of years elapsed since the award, with T = 0 representing the date of the award. T* is not an estimate of when payment will occur. It is the turning point at which the award's discounted present value reaches its maximum. Before that point, additional award interest more than offsets the effect of discounting; after it, discounting becomes the stronger effect.
Consider an award indexed to 100 on the date it is made, accruing simple interest at ra = 12%. The table shows its discounted present value when discount rates of 7%, 10% and 25% are applied:
| Years elapsed | Discounted value at rd = 7% | Discounted value at rd = 10% | Discounted value at rd = 25% |
|---|---|---|---|
| 0 | 100.0 | 100.0 | 100.0 |
| 2 | 108.3 | 102.5 | 79.4 |
| 4 | 112.9 | 101.1 | 60.6 |
| 6 | 114.6 | 97.1 | 45.1 |
| 8 | 114.1 | 91.4 | 32.9 |
| 10 | 111.8 | 84.8 | 23.6 |
| 12 | 108.3 | 77.7 | 16.8 |
Illustrative parameters. Not calibrated to any actual claim.
For a discount rate of 7%:
T* = 1 / ln(1.07) − 1 / 0.12 ≈ 6.4 years
The 6.4 years are measured from the date of the award. The discounted value rises until approximately that point, reaching a maximum of about 114.7. Until then, the additional interest accrued more than offsets discounting. After 6.4 years, the discounted value begins to fall, although it remains above its award-date value even after twelve years.
At a discount rate of 10%, T* is approximately 2.2 years. The discounted value reaches a much earlier and smaller maximum of about 102.5 before beginning to decline. By year twelve, it has fallen to approximately 77.7.
At a discount rate of 25%, T* is approximately −3.9 years. The negative result means that no turning point exists after the award date: the discounted value falls from the outset because discounting outweighs the additional interest throughout the period shown.
The same award, interest rate and payment timetable can therefore produce materially different valuation paths depending on the discount rate applied. Below a sufficiently high discount rate, award interest may initially offset — or exceed — the erosion of present value. Above that rate, elapsed time reduces present value from the beginning.
"What is this claim worth?" is an incomplete question
Present value is not a property of a claim alone. It is a property of a claim and the assumptions used to value it. Our note on contingent claim valuation observes that different parties can reach different present values on identical recovery assumptions; the turning point is one mechanism by which that happens.
It follows that a disagreement about a timetable need not be a disagreement about the claim. Two parties working from the same facts, the same merits assessment and the same award interest rate can reach opposite conclusions about whether an additional six months helps or hurts, purely because they discount at different rates. Suppose, illustratively, that two parties apply discount rates of 9% and 19% to an award carrying simple interest at 12%. T* is approximately 3.3 years for the first and −2.6 years for the second. A further year of delay from the award date moves the first valuation up by roughly 3% and the second down by roughly 6%. Illustrative parameters. Not calibrated to any actual claim. The turning-point calculation makes the source of that difference visible.
The effect of delay depends on the discount rate applied
The model is deliberately simple. It assumes a fixed rate of simple interest, a constant discount rate, an assumed payment date and no further costs. In practice, costs may continue to accumulate, the interest entitlement depends on the terms of the award, and the payment date remains uncertain. Simulation can incorporate that uncertainty, producing a distribution of possible values rather than a single estimate.
But the central point survives those simplifications: delay has no single effect on present value. It can increase or reduce the value of the same award, depending on the relationship between the award interest rate and the discount rate applied. A valuation therefore cannot be properly understood without knowing the rate behind it. When two parties value the same interest-bearing award differently, the disagreement may lie not in their view of the merits, but in how each values time.
5 Rivers Capital publishes research on the valuation of legal claims as an asset class. This note is analytical and is not investment advice, legal advice, or an offer or solicitation in respect of any security or fund interest. Any figures shown are illustrative and are not calibrated to any actual claim.
