The Cash Discipline

Gold Kept Its Value and Little Else

The empirical case against non-distributing stores of value is now unusually well documented, and it comes from a source with no obvious axe to grind.

The UBS Global Investment Returns Yearbook 2026, published on 3 March 2026 and compiled by Professor Paul Marsh and Dr Mike Staunton of London Business School with Professor Elroy Dimson of Cambridge, draws on 126 years of data covering equities, bonds, bills, currencies and, for the first time in this form, gold, across 35 markets. Its finding on gold is precise. Over the very long run gold has preserved purchasing power, with the real dollar gold price rising 5.2-fold since 1900, equivalent to an annualised real return of 1.3%. The Yearbook's framing of gold's role is that it is nuanced as an inflation hedge, noting that gold delivered negative returns in 13 of the 28 years in which inflation exceeded 3%.

Set that against the same dataset's finding for equities. In real terms, after inflation, equities delivered 6.6% a year from 1900 to 2025, against 1.6% for bonds. In nominal terms, one US dollar invested in equities in 1900 became USD 124,854 by the end of 2025, against USD 284 in long bonds and USD 69 in Treasury bills. Since 1900, equities have outperformed bonds, bills and inflation in every country with a continuous investment history.

The gap between 1.3% and 6.6% compounded over 126 years is not a matter of degree. It is the difference between preserving capital and creating it. And the mechanism behind the gap is not complicated. An equity is a claim on a stream of cash produced by an operating business. Gold is a claim on nothing. Its return is entirely the difference between the price someone paid and the price someone else will pay, and over the very long run that difference has, in real terms, been small.

This is not an argument that gold is useless. The Yearbook is explicit that gold has sometimes been effective as an equity market hedge, and the family office allocation of 2% is consistent with holding it as insurance rather than as an investment. Insurance premiums are supposed to cost something. The error is not owning gold; the error is expecting gold to compound.

The general principle extends beyond gold to every asset whose entire return depends on resale. Fine art, classic cars, collectible watches, vintage wine and the rest share the structure: the holder pays to store, insure and maintain the object, receives nothing in the interim, and must find a buyer at a higher real price to earn anything at all. The 1% allocation family offices give to art and antiques is a reasonable assessment of that structure. It is enough to own things one likes. It is not enough to matter.

Distribution as a Governance Mechanism

The conventional explanation for why sophisticated capital prefers cash-producing assets is that income is a component of total return and a reliable one. That is true and it is the least interesting reason.

The more important reason is that a distribution is a governance mechanism. It does three things that no valuation can do.

First, it forces the manager to realise. An asset that must distribute cash quarterly cannot be run indefinitely on projections. Someone has to collect a receivable, bank it and pay it out. That requirement disciplines everything upstream of it: the tenant has to pay rent, the borrower has to service the loan, the charterer has to settle the invoice. A strategy whose entire return is deferred to an exit has no such forcing function, and the forcing function is exactly what is absent in the ten-year private fund. The CFA Institute's Sebastien Canderle cited an analysis by the secondaries adviser Palico of 200 private equity funds in which more than 85% failed to return investors' capital within the ten-year term. When realisation is optional, realisation is postponed.

Second, it constrains valuation fiction. A manager marking an asset upward while distributing nothing is making a claim that cannot be tested until exit. A manager distributing cash is producing evidence, every quarter, that is independent of their own opinion. The two can diverge, and the divergence is itself information: a rising mark and a falling distribution is a signal, and it is a signal that only exists in a structure where distributions happen.

Third, it changes the investor's position over time. Capital returned is capital no longer at risk in that asset. An investor receiving distributions has a declining exposure to a single manager's judgement with every payment, and can redeploy at prevailing market rates rather than at the rates that prevailed when the commitment was signed. In an environment where the ten-year real risk-free yield has moved from 1.94% at the start of January 2026 to 2.41% by mid-August, according to the United States Treasury's daily par real yield curve, the ability to redeploy is worth something concrete.

The behaviour of family offices is consistent with this. Knight Frank's Family Office Survey 2026, drawn from interviews with more than 40 family offices across London, New York, Dubai, Singapore and Hong Kong, reports that operational real estate sectors have emerged as attractive opportunities, singling out data centres, student accommodation, healthcare and logistics as combining stable income with long-term growth potential. Note the word operational. These are not land banks. They are businesses with tenants, occupancy, operating costs and a cash cycle.

The same survey records that direct ownership remains attractive because it allows families to shape development strategies, manage risk, and capture the full upside rather than sharing returns through fund structures. That preference is worth stating plainly, because it cuts against the fund model rather than for it. Families who can operate an asset directly frequently do, and the fund layer is a compromise made by investors who cannot.

What a Mark Cannot Tell an Investor

The case for distribution is strongest when set against what the alternative source of information actually is, which is a valuation produced by a process the investor cannot observe.

The Financial Stability Board's report on vulnerabilities in private credit, published on 6 May 2026, is the clearest public description available of how such valuations are produced. Private credit valuations, it records, are typically based on a discounted cash flow or yield analysis for performing loans, supplemented by public market proxies such as leveraged loan indices or high-yield bond markets used as benchmarks for comparable peer companies. They are calibrated at origination and adjusted over time for changes in credit quality, borrower performance and market conditions.

Then come the qualifications. Valuations are updated infrequently, often on a quarterly basis, which the FSB describes as possibly adequate in normal market conditions but less so under stress. The illiquidity of the assets and the cost of valuing them make more frequent revaluation impractical. Investors may have only limited access to granular information on loans they are not invested in, which the FSB notes may affect pricing efficiency and increase the dispersion of valuations and related marks.

The report then addresses incentives directly, and this is the passage that matters most for anyone relying on a reported net asset value. Perceived or actual stale valuations, it states, may create a first-mover incentive during stress events, leading investors to exit a fund before asset values are potentially marked down. Managers, it continues, may have potential incentives to manage valuations of their funds in a way that minimises the appearance of volatility, such as by delaying or spreading out the impact of negative shocks that could reduce asset values. On the underlying problem its conclusion is unequivocal: while robust governance arrangements may help address some of these concerns, the relative opacity of private credit cannot be easily resolved.

The European Central Bank's Financial Stability Review of May 2026 supplied the live demonstration. Semi-liquid private credit vehicles in the United States were hit by sizeable redemption requests from the beginning of 2026. Some funds met investor requests in full; others capped redemptions at a specified share of fund assets in line with contractual agreements. The ECB's own reading is that these outflows illustrate how a deterioration in risk sentiment can spur investors to withdraw their capital from funds offering redemptions at a regular frequency, despite their portfolio holdings being less liquid.

Every one of these problems is a property of a mark. None of them is a property of a payment. A quarterly valuation can be late, anchored, discretionary and smoothed. A quarterly bank transfer either arrives or does not, and the investor knows within a day. That asymmetry is the real argument for distributing assets, and it is an argument about information rather than about return.

How a Yield Is Manufactured

Everything above would justify screening investments on distribution rate. That would be a mistake, and it is worth being specific about why, because an investor who selects on headline yield is the easiest investor in the market to mislead.

A distribution can be funded from four sources. Only one of them is the one the investor thinks they are buying.

The first is operating cash flow generated by the asset in the period. This is the real thing.

The second is return of the investor's own capital. A vehicle that raises EUR 100 million and pays out EUR 8 million a year can sustain a headline 8% distribution rate for a considerable period without earning anything at all, simply by paying investors with their own subscriptions and those of later investors. The distribution rate looks identical on a factsheet. Only the reconciliation between distributions and cash generated from operations distinguishes them, and that reconciliation is frequently not published.

The third is borrowing. A vehicle can draw on a credit facility to smooth or support distributions through a weak period. Done transparently and for a genuine timing mismatch this is ordinary treasury management. Done to preserve an advertised rate through a structural deterioration it is a transfer from future investors to present ones.

The fourth is deferred income booked as though it had been received. The FSB report documents this mechanism precisely in private credit, where payment-in-kind arrangements allow borrowers to defer cash interest by adding it to loan principal. The FSB distinguishes payment-in-kind notes, where deferred interest is an embedded feature drawable from origination, from payment-in-kind toggles, which allow a borrower to elect deferral later in the loan term and which the report notes some market participants colloquially describe as bad payment-in-kind. Its findings are specific: payment-in-kind is used in approximately 12% of loans, with toggles accounting for about half of those cases, and the use of both has risen significantly since 2022, coinciding with rising rates, without showing signs of decline. A study of loans in United States business development companies cited by the FSB found that the use of payment-in-kind toggles is associated with a one to two percentage point increase in the likelihood of a loan becoming delinquent in the following quarter, against an unconditional probability of 3%, often accompanied by a deterioration in the loan's valuation.

Read that mechanism carefully in the context of a yield. Interest that a borrower cannot pay in cash is added to the loan, recognised as income by the lender, and can support a distribution to the lender's investors. The distribution is real cash leaving the fund. The income supporting it is an accounting entry describing a payment that did not happen. The FSB's summary of the broader trend is that some private credit borrowers appear to be relying more on payment-in-kind loans, which can also signal deteriorating credit conditions.

The ECB provides the corroborating stress indicator. The ability of private credit-backed firms in the euro area to service interest payments from operating cash flows, it reports, has deteriorated in recent years, a trend also present in leveraged loan and high-yield bond markets but absent among firms relying on bank loans.

An investor screening on headline distribution rate cannot distinguish between any of these four cases. They all produce the same number.

Separating a Real Distribution From a Funded One

The distinction is nevertheless knowable in advance, and the questions that make it knowable are short, specific and answerable.

The first question is the reconciliation. For each distribution period, what was distributed, and what was cash generated from operating activities in that same period. If distributions exceed operating cash flow, the difference came from somewhere, and the manager should be able to say from where and why. A structure that cannot produce this reconciliation on request has not been built to be examined. A structure that produces it and shows a persistent shortfall is returning capital, which may be entirely legitimate but is not a yield.

The second question is the contractual basis. Is the payment obligation contractual, and if so, who owes it, to whom, and what is their standing to pay. A commitment given by a well-capitalised operator to a fund is a materially different instrument from a projection produced by the fund's own model. The former can be assessed by examining the obligor. The latter can only be assessed by trusting the assumptions.

The third question concerns deferral. Does any part of reported income consist of interest, rent or fees that were accrued rather than received in cash. This is the payment-in-kind question generalised, and it applies well beyond credit. Rent-free periods, deferred charter fees, capitalised management charges and receivables of increasing age all have the same effect: income today, cash later or never.

The fourth question is coverage under stress. At what level of occupancy, utilisation or rate does the distribution stop being covered by operations. The answer is a single number and every competent operator knows it. It is rarely in the marketing document because it is the number that defines the downside.

The fifth question is who bears the operating cost base and on what terms. In an operating real asset the cost line is not a footnote; it is the difference between gross revenue and anything distributable. Whether fuel, maintenance, crew, insurance and berthing sit with the owner or pass to the customer changes the risk profile completely while leaving the headline yield unchanged.

The sixth question is the seniority of the investor's claim to that cash. A distribution that ranks behind a lender, a manager's fee and a preferred class of interests is a residual, and residuals behave very differently from contracted payments in a bad year.

None of these six requires access to proprietary information. All of them are answerable from an offering document and a set of management accounts. An investor who asks them will fail to be misled by a manufactured yield, and will also, quite often, find that a lower headline rate covered two times by operating cash flow is worth more than a higher one covered once.

Cash Flow Against a Real Hurdle

The final discipline is the one that applies to every proposition in 2026 regardless of its structure.

On 18 August 2026 the United States Treasury's daily par real yield curve put the ten-year real yield at 2.41%, the five-year at 2.10% and the thirty-year at 3.03%. At the start of the year the ten-year real yield was 1.94%. An investor can buy a real return of roughly 2.4% a year for a decade, with daily liquidity, no operating cost, no manager, no gate and no valuation discretion.

That number is the floor under every cash-flowing asset argument, including this one. A distributing asset must clear it after operating costs, after fees and after whatever discount the investor applies for illiquidity and for concentration in a single operator. In 2021, when real yields were negative, any positive distribution looked attractive on its own terms. That comparison is no longer available, and a great deal of income-focused marketing has not adjusted to the fact.

The honest reading of the evidence in this article is therefore narrower than the headline suggests. Cash-producing assets are preferable to non-producing stores of value not primarily because they return more, though the 6.6% against 1.3% comparison over 126 years is not nothing, but because they generate verifiable information while they are held. That informational advantage is real, it is durable, and it survives the counterargument. What does not survive is the idea that a distribution rate is itself evidence of anything. The FSB's payment-in-kind findings are a live demonstration that income can be recognised without cash changing hands, in a market that institutional investors entered specifically for its yield.

HelmShare's structure is built around cash generated by an operating fleet rather than around an exit valuation, and is offered here as one worked example rather than as the answer to the question. HelmShare Prime Fund, L.P. is a Cayman Islands exempted limited partnership issuing limited partnership interests. Its general partner and investment manager is HelmShare LLC, regulated by the Dubai Financial Services Authority in the DIFC as a Category 3C firm. Interests are offered under Regulation S outside the United States to professional, qualified and high net worth investors in the EU and EEA, the United Kingdom and the Gulf, and US persons are excluded. The fleet is operated by Navigare, which has given a contractual 8% yield commitment to the Fund. That commitment runs to the Fund and is not a promise to any individual investor. Returns are targeted rather than assured, capital is at risk, and the six questions set out above apply to HelmShare exactly as they apply to anything else.

The reason sophisticated capital gravitates toward assets that distribute is not that distribution is proof of quality. It is that distribution is falsifiable. A mark is an opinion that becomes true or false only at exit, which may be years after the investor needed to know. A payment is a fact that arrives on a schedule, and an investor who is receiving cash is being told something every quarter that an investor holding a valuation is not being told at all. That is the whole of the advantage, and it is enough.