When the Target Return Holds Still and the Risk Moves: The Institutional Mechanics of Reach for Yield

The target return entered into the minutes of an investment committee is, in most institutions, inherited from the previous year's minutes rather than derived from the current year's market conditions; once the agenda opens, what gets debated is not whether that figure still holds but which asset might deliver it. Even in a period during which the reference rate, credit spreads, and funding costs have all compressed materially over twelve months, the number itself is seldom tabled as a discussion item, because it belongs not to the committee but to a commitment that preceded it — a fund document, an actuarial assumption, a group budget, a representation made to shareholders. The observable behaviour is a migration: discussion moves from the reasonableness of the target to the search for a route to it, and this migration is nowhere recorded as an independent decision. The language of approval thins in parallel, so that formulations such as "same return, slightly longer tenor," "one step further down the capital structure," or "exit contingent on a narrower window" pass through the sentence as qualifiers when each of them, taken on its own terms, constitutes a risk disclosure.

The same pattern recurs, in different dress, at the corporate treasury desk, in the office of a group finance director, and in a project finance credit committee. On the treasury side, surplus cash migrates from deposits into structured products and from there into instruments whose maturity extends beyond the liquidity horizon the balance sheet actually requires, with each step supported by a rationale that is defensible in isolation; in the credit committee, accepting mezzanine paper in place of senior debt, or an unrated counterparty with an apparently robust security package in place of a rated one, is defensible in precisely the same isolated way. On the project side, the same movement appears as a shift from an operating asset to one that has not yet reached commercial operation, or from a portfolio underpinned by long-tenor PPAs to one carrying partial merchant exposure. None of these decisions is erroneous standing alone; what is defective is that their sum is never consolidated in any document as a single movement.

The behaviour has a name — reach for yield, sometimes rendered as search for yield — and it denotes the tendency, under a fixed return target, to close the residual gap by conceding on the quality of risk. The mechanism operates on a specific asymmetry: return is nominal, one-dimensional, measurable and reportable on a quarterly cadence, whereas risk is multidimensional, probabilistic, and typically visible only after a lag. To the extent that a committee's performance is assessed on what can be measured, the concession is always made on the side that resists measurement, and within the measurement period that choice carries no visible cost. The persistence of the tendency follows from this: the shortcut is priced not when it is taken but long after the conditions that justified it have changed.

Under certain conditions the tendency is not an error at all but a direct function of the balance sheet. For an institution whose liabilities are fixed in nominal terms — a pension obligation, an insurance reserve, a debt service schedule contractually bound in both currency and rate — the pursuit of return is not a preference but a requirement imposed by the arithmetic of the liability side, and in a market where spreads have genuinely repriced, moving into a higher-yielding asset does not necessarily mean taking incremental risk. The distinction lies elsewhere: whether spread compression reflects a reduction in underlying risk or an increase in capital chasing the same pool of assets. In the first case, holding the target is reasonable; in the second, holding the target means accepting structurally weaker documentation for the same headline number. A question capable of separating these two states rarely appears anywhere on the committee agenda.

The surfaces on which a concession can be made are finite, and they typically run along five lines: extending tenor, accepting lower credit quality, relinquishing liquidity, moving one step down the capital structure, and stepping backward in the maturity stage of the asset itself. What these five share is that none of them registers on the return line; all five produce the same headline figure, yet their behaviour under a loss scenario differs entirely. Extended tenor generates repricing risk, forgone liquidity generates timing risk, and subordination generates a direct recovery risk, and when these accumulate the resulting profile may prove several times more sensitive than the profile the committee believed it was approving. A technical reading of the structure establishes the point plainly: two transactions carrying an identical nominal return can behave an order of magnitude apart under stress.

The institutional cost first appears not in price but in documentation. In periods when the search for return intensifies, covenant packages thin, DSCR test thresholds loosen, the funding schedule for reserve accounts is deferred until after closing, security collapses into a pledge over the shares of the project company, the survival period for representations and warranties shortens, and the escrow proportion falls. None of these items generates a line in the internal rate of return calculation and none therefore appears in the investment memorandum, yet they are precisely the terms that determine the recovery rate in a loss scenario. The most direct route to reading the actual risk profile of a portfolio runs not through the distribution of returns but through a comparison of the covenant headings in agreements signed over the last twelve months against those signed in the twelve months preceding them.

The moment at which the cost crystallises is generally not first draw but the refinancing window. A capital structure assembled on an assumption of low funding cost, arriving at maturity in a market where that assumption no longer holds, leaves the sponsor with three routes — injecting additional equity, selling the asset at a discount, or returning to the negotiating table with the lender — and all three produce an outcome below the return originally modelled. For this reason the cost of reach for yield is usually booked not as a loss but as a delay and a dilution, and accounting does not present it in any single line item. Where tail-period obligations happen to fall in the same window — replacement capital expenditure, decommissioning and restoration undertakings, long-term maintenance agreements — the effects compound rather than arrive sequentially.

The second cost is concentration. Assets accumulated within a single funding window may appear well dispersed by sector — an industrial facility on one side, a data centre on another, a storage project on a third — while sharing a single assumption: that capital will remain cheap and available. This shared assumption is nowhere visible in the portfolio table, because the table sorts assets by sector, geography, and technology rather than by funding assumption. When a counterparty conducting due diligence in a sale or partnership process identifies the concentration, the consequence typically registers not in the valuation of individual assets but in a portfolio-level discount, an expansion of conditions precedent, or the attachment of a portion of consideration to an earn-out. What determines valuation is not the historical return of the portfolio but whether that return can be shown to be repeatable once the founding assumption changes.

This tendency cannot be managed through individual awareness, because what produces it is not the judgement of the decision-maker but the manner in which the target has been defined; the intervention must therefore be structural in kind. Four applicable components separate out: first, defining the target return as a spread over the risk-free rate rather than as a nominal figure, so that the target reprices automatically as the rate environment moves; second, tying approval authority to structural attributes — tenor, seniority, liquidity, maturity stage — rather than to the magnitude of return, such that any structural migration beyond a defined threshold requires its own approval; third, maintaining a reasoned record of declined transactions, since the drift of a committee's risk threshold is legible only through what it refuses; fourth, instituting a regular review cadence in which the portfolio is re-sliced by funding assumption rather than by asset class.

BEIREK's intervention in this problem is directed not at the investment decision itself but at the recording architecture within which the decision is constructed. On the projects we manage, the decision record opens at the moment of proposal rather than at the moment of approval; each transaction file carries a structural profile page that decomposes which of the five lines the return figure has been sourced from, and that page presents, side by side, the structure through which the same target return was produced in the preceding period. On the contractual side, the position of the covenant package, the reserve account calibration, and the scope of security relative to comparable transactions within the portfolio is reported on a single page before closing, so that the concession made in the documentation reaches the table simultaneously with the economic headline the committee is already reviewing. The purpose is not to slow the transaction but to render the line along which the concession has been made visible at the same moment as the decision.

The second line of intervention is cadence. In the quarterly portfolio review we operate, assets are sliced by funding assumption and refinancing calendar rather than by sector, and the aggregate size of obligations maturing within the same window is consolidated into a single table; that table is the instrument through which the concentration produced by individually reasonable decisions becomes visible in aggregate. The stakeholder pre-mortem we run ahead of each material transaction attaches a single question to a formal role: if this transaction underperforms three years from now, which of the assumptions currently sitting on the table is most likely to have been the source. Whether that role sits with someone independent of the sponsoring team determines the outcome; where a team is asked to generate counter-arguments against its own proposal, the arguments produced are predictably weak.

The risk appetite of an institution resides not in its written policy document but in the way its target return is defined; a fixed nominal target functions as a mechanism that silently rewrites risk appetite each time the rate environment moves, and that rewriting is never put to a vote in any boardroom. Where the returns on transactions being approved today match those of their counterparts from three years ago, the question worth asking is what accounts for the stability.