Loss Experience in GCC Private Credit: Is the Yield Real?

Introduction

There is a question that sits beneath every conversation about Gulf private credit, often unspoken because it sounds naive, yet more important than any other: is the yield real? A coupon of fifteen or eighteen per cent is easy to quote and easy to admire, but the return an investor keeps is the coupon less the losses, costs and fees that the full life of the book imposes. If losses are modest, the high coupon translates into a high net yield and the investment case is made; if losses are large, the coupon is merely the compensation for a loss that arrives, and the net yield collapses. The headline coupon, in other words, is a gross figure that says nothing definitive about the return until it is tested against loss.

This anxiety is rational and should not be dismissed. New, high-yielding markets have a long history of disappointing investors who confused the coupon with the return, deploying capital on the strength of a headline yield only to watch a credit cycle reveal that the yield was compensation for risk all along. A disciplined investor approaching Gulf private credit is right to ask whether the Gulf will prove the same, and to demand evidence rather than assurance. The purpose of this paper is to supply a framework for generating and interrogating that evidence, by modelling loss experience explicitly and testing the yield against it.

The framework rests on a distinction that the earlier papers developed and this one applies: the gross coupon is mostly compensation for illiquidity, complexity and the supply of scarce capital, with a smaller component compensating for expected credit loss. If that decomposition is correct, then the credit-loss component, the part that realised losses will consume, is a minority of the coupon, and the remainder is a buffer that losses must exhaust before the net yield is impaired. The question whether the yield is real therefore becomes a question about the size of that buffer relative to the losses a full cycle can impose: if the buffer is large relative to plausible losses, the yield is real and durable; if it is thin, the yield is fragile.

The paper approaches this through modelling rather than through a proprietary loss dataset, for the honest reason that Gulf private credit is too young to have generated a long, public record of realised losses through a full cycle. This is a genuine limitation, and the paper neither hides it nor pretends to data it does not have. Instead it models loss experience transparently, calibrated to the general loss experience of comparable private-credit markets and to the structure of the Gulf market, and it stresses the model hard, so that the conclusions rest not on optimistic assumptions but on how the yield behaves across a wide range of loss outcomes including severe ones. The aim is not to predict Gulf losses but to establish how much loss the yield can withstand.

Results And Discussion

This section presents the modelled loss experience and the test of the yield: cumulative loss curves, default and loss by tier, the net-yield bridge, the comparison with developed markets, and the size of the yield buffer.

Cumulative Loss Curves by Vintage

Figure 1 shows modelled cumulative loss curves for three vintages: a benign vintage originated and run in good conditions, an average vintage, and a downturn vintage that meets a property downturn during its life. The curves diverge sharply. The benign vintage accumulates modest losses; the average vintage more; the downturn vintage substantially more, with losses accelerating in the years the downturn bites. The figure makes Proposition 3 concrete: the loss a book experiences depends heavily on whether it meets a downturn, and a book judged only by its benign early years will understate the loss a full cycle can impose.

Loss experience depends heavily on whether a vintage meets a downturn; benign early years understate full-cycle loss.

The vintage analysis carries an investment lesson about timing and patience. A book originated near the top of a cycle and tested by the subsequent downturn will show the downturn curve; one originated after a correction, into a recovering market, will show the benign curve. The investor cannot control which vintage it is buying into with certainty, but it can recognise where in the cycle it is originating, lend more cautiously near the top and more confidently after a correction, and judge a manager's loss experience in the light of the vintages it has actually lived through rather than crediting benign early returns as proof of durable skill.

Default and Loss by Tier

Figure 2 shows modelled annual default and loss rates across the five underwriting tiers. Both rise with the tier, and the loss rate rises faster than the default rate, because the higher tiers combine higher default probability with higher loss given default, the two reinforcing each other. The figure quantifies Proposition 4: the loss a book experiences depends on its tier composition, and a book concentrated in the high tiers will experience materially higher losses than one tilted toward the low tiers, even at the same average coupon. The tier composition is therefore as important to loss experience as the average coupon is to gross income.

Figure 2. Default and Loss Rate by Underwriting Tier

Both rise with the tier, and the loss rate rises faster, because high tiers combine higher default and higher loss given default.

Table 2 sets out the tier-by-tier default and loss detail and the resulting expected loss, the input to the net-yield bridge. The table shows that even the higher tiers have expected losses well below their gross coupons, so that a buffer exists at every tier; but the buffer narrows as the tier rises, because the loss grows faster than the incremental coupon. This is the quantitative content of Proposition 4: the yield is real at every tier, but its buffer, and hence its durability under stress, is largest at the senior end and thinnest at the subordinated end.

Table 2. Modelled Default, Loss and Buffer by Tier (Illustrative)

Illustrative; buffer is gross coupon less expected loss, before costs and fees. Buffer in absolute terms is large at all tiers but proportionally thinner at high tiers.

The Net-Yield Bridge

Figure 3 presents the net-yield bridge for the blended book, moving from the gross coupon through the default loss, the recovery that offsets part of it, servicing and currency costs and fund fees, to the net yield. The bridge shows that, under base-case losses, the net yield remains firmly in double digits, well below the gross coupon but well above developed-market alternatives. The default loss, before recovery, is the second-largest deduction after fees, but the recovery, the subject of the security and enforcement paper, offsets a substantial part of it, illustrating how the recovery engineering of that paper feeds directly into the net yield of this one.

Figure 3. From Gross Coupon to Net Yield After Losses

Under base-case losses the net yield remains double-digit; recovery offsets much of the default loss, fees are the largest deduction.

The bridge is the operational answer to the paper's title question under base-case conditions: the yield is real, a net figure several points below the headline but substantial, and the deductions that separate gross from net are dominated not by losses but by fees, with losses, after recovery, a smaller drag. This reinforces a theme of the series: for well-underwritten, well-secured Gulf credit, the credit-loss component of the coupon is a minority of the total, so realised losses, while real, do not consume the yield under normal conditions.

Comparison with Developed-Market Loss Experience

Figure 4 compares modelled Gulf loss rates with the loss experience of developed-market private credit at matched risk tiers. The modelled Gulf loss rates are somewhat higher, reflecting the younger market and the property-cycle concentration, but not dramatically so, because the strong collateral and recovery engineering examined in the earlier papers limit loss given default. The comparison matters because the Gulf coupon is substantially higher than the developed-market coupon at the same tier, while the modelled loss is only modestly higher; the gap between the larger coupon premium and the smaller loss premium is the extra buffer that makes the Gulf net yield attractive.

Figure 4. Modelled GCC versus Developed-Market Loss Rates

Modelled Gulf losses are modestly higher than developed-market losses; the coupon premium is far larger than the loss premium.

Implementation Considerations

Turning the evidence into practice requires attention to loss tracking, tier discipline, diversification, manager assessment and the honest treatment of an untested record.

Tracking and Building Loss Data

Because the market lacks a long public loss record, the investor's own loss data are uniquely valuable, and the discipline of tracking them rigorously is a priority. Every default, recovery and loss should be recorded against the original underwriting assessment, so that, over time, the investor accumulates a proprietary dataset that reveals its actual loss experience and calibrates its future modelling. A lender or allocator that tracks losses rigorously will, after a few years, hold evidence that no newcomer can replicate, and that evidence is both a source of edge and the only true answer to the question whether the yield is real for its particular book.

Tier Discipline

Because the yield buffer narrows as the tier rises, tier discipline, controlling the proportion of the book in the high tiers, is the principal lever over the durability of the net yield. A book tilted toward the senior tiers sacrifices some gross coupon for a larger buffer and a net yield that survives stress; a book that maximises coupon by concentrating in the high tiers earns more in benign conditions and risks a negative net yield in stress. The investor should set and hold a tier allocation that delivers an acceptable net yield under stress, not merely an attractive one under base-case conditions.

Diversification and Its Limits

Diversification across sponsor, asset type, sub-region and vintage widens the effective buffer by reducing idiosyncratic loss, and the investor should pursue it deliberately. But the scenario and vintage analysis showed that diversification cannot eliminate the systematic property-cycle exposure that all the loans share, so the investor must size the overall allocation on the assumption that the systematic factor will eventually bind. Diversification reduces the loss in normal conditions and the idiosyncratic part of the tail; it does not protect against the correlated downturn, which only the buffer and conservative sizing can absorb.

Assessing a Manager's Loss Experience

For the allocator, assessing a manager's loss experience is delicate precisely because the market is young and most managers' records are benign and untested. The allocator should ask not only what losses the manager has experienced but through what vintages, distinguishing a record built entirely in benign conditions from one tested by a downturn. It should examine the manager's tier composition, its underwriting and recovery discipline, and its loss-tracking rigour, treating these as predictors of future loss more reliable than a short, benign track record. A manager with a disciplined process and a modest but tested record is preferable to one with a spectacular but untested one.

The Honest Treatment of an Untested Record

Both managers and allocators should treat an untested record honestly, neither dismissing it as worthless nor crediting it as proof of durable skill. A few years of low losses in benign conditions is genuine evidence of competent origination and underwriting, but it is not evidence that the yield survives a downturn, because the downturn has not occurred. The honest position is to present the record for what it is, to model the loss a downturn would impose, and to test the yield against that modelled stress, exactly as this paper does. An investor who confuses a benign record with a stress-tested one is making the very error the paper warns against.

Pricing the Buffer into the Allocation Decision

The yield buffer should be priced into the allocation decision as a margin of safety. An allocator should size its Gulf credit allocation, and choose its tier composition, so that the net yield remains acceptable under a stress scenario it regards as plausible, treating the base-case net yield as the expected outcome and the stressed net yield as the floor it is willing to tolerate. This converts the abstract question whether the yield is real into a concrete allocation discipline: allocate to the point where the stressed net yield is still acceptable, and no further. The buffer, properly priced, becomes the governor of the allocation size.

Communicating Loss Risk to the Committee

Concluding Comments

Before drawing the threads together, it is worth restating the question in the plainest terms, because it is the question every prospective investor in this market silently asks. Strip away the modelling and the scenarios and the issue is simple: if I lend into the Gulf at these coupons, and a normal share of my borrowers fail, will I still make money? The whole apparatus of this paper exists to answer that plain question with evidence rather than assertion, and the evidence says yes for the disciplined lender and no for the careless one. Everything else is detail in service of that answer.

This paper has addressed the question on which the Gulf private-credit case ultimately turns: once realised defaults and losses are taken into account, is the yield real? The answer, developed through transparent loss modelling and demanding stress testing, is that for well-underwritten, well-secured, diversified Gulf credit the yield is real, a net figure with a substantial buffer that survives all but severe stress, while for poorly-underwritten, weakly-secured, concentrated books the yield is fragile and can be consumed by loss. The yield is real for those who earn it through discipline and an illusion for those who chase it without.

The practical implications follow. The gross coupon provides a large buffer over expected loss, but the buffer narrows as the tier rises, so tier discipline is the principal lever over the durability of the net yield. Losses are cyclical and correlated, so the benign experience of a young, untested book understates the loss a full cycle can impose, and the yield must be tested against the downturn rather than the expansion. Diversification widens the buffer but cannot eliminate the systematic property-cycle exposure, which only the buffer and conservative sizing can absorb. And the investor's own loss data, tracked rigorously, are the only true answer to whether the yield is real for its particular book.

The argument is properly and importantly qualified by the data limitation at its heart: Gulf private credit has not yet experienced a full credit cycle, so the loss experience modelled here is calibrated to comparable markets and to market structure rather than measured directly. The paper does not hide this; it responds to it by modelling transparently, calibrating conservatively and stressing hard, so that the conclusion rests not on optimistic loss assumptions but on how much loss the yield can withstand. The honest claim is not that Gulf losses will be low but that the yield can absorb losses materially higher than the base case before it ceases to be attractive, for a well-constructed book.