So, can reserves separate market risk from immediate liquidity?

The answer at the center of this series is yes, but only in a narrow and useful sense. A Synthetic Reserve Deposit (SRD) can separate the time at which a reserve holder needs foreign-currency cash from the time at which market-risky assets must be sold. It cannot separate the SRD claim from the market value of those assets. This final judgment connects the floating-value balance sheet, the redemption and borrowing routes, the calibrated portfolio results, and the stress and scale exercises so that the reader can tell a genuine liquidity improvement from a disguised guarantee.

What can be separated—and what cannot

The cleanest way to judge the proposal is to ask what happens after a shock, not what the portfolio earns in an average month. Suppose interest rates rise sharply and the diversified assets inside the pool lose market value. The SRD unit value falls because the claim is floating. That loss is not an accounting inconvenience to be smoothed away; it is the market outcome borne by the participants. The pool has not separated them from market risk.

What the pool may change is the next decision. A participant that needs cash can redeem units at the pool’s contemporaneous executable value. If the settlement buffer is sufficient, the pool can pay without selling investment assets immediately. If the buffer is insufficient, the pool sells assets and the current execution prices determine the amount paid. The participant receives cash, but a loss may be realized at that point. The redemption route separates the request for cash from the participant’s need to manage the pool’s entire portfolio, not from the value of the assets being sold.

A participant with a temporary need may instead pledge units for a short-term advance. The claim remains outstanding, so a later recovery in the underlying assets remains economically relevant to the participant. But the participant now owes repayment, the units are subject to a haircut and margining, and the pledged portion is encumbered. If the assets continue to fall, the participant’s exposure does not disappear; the collateral gap may require more collateral, partial repayment, substitution, or eventually liquidation. The borrowing route separates the timing of cash from the timing of an outright sale, but it does not turn the pledged claim into stable money.

This distinction can be expressed as three different promises. The first is a valuation promise: the claim will be priced at a rule that reflects what the pool can actually realize. The second is a funding promise: a participant may obtain short-term cash through a buffer, a redemption, or—if available—a secured advance. The third is a reserve-status promise: the claim will be recognized as an external foreign-currency asset controlled by the monetary authority and usable within the required time. The SRD architecture addresses the first two through its operating rules, but the third depends on law, accounting, governance, and institutional acceptance.

Confusing those promises is how a liquidity design becomes a false safety claim. A large buffer may make ordinary settlement convenient, but it cannot guarantee that all participants can withdraw at a protected price. A public liquidity provider may make a temporary advance more credible, but it is not supposed to guarantee unit value or absorb ordinary portfolio losses. A portfolio may have a lower bad-state loading than equities, but it can still lose value after a rate jump or an equity decline. The proposal is coherent only if each benefit is attached to the mechanism that produces it and each risk remains attached to the party that bears it.

The same logic explains why contemporaneous executable value is more important than the word “deposit.” If redeemers could receive a stale pre-sale value while remaining holders absorbed the difference, the pool would be transferring market losses between participants. If a public backstop guaranteed principal, the arrangement would be transferring those losses to the provider. The benchmark design avoids both promises. Redeemers and remaining holders are valued on the same basis, and the collateralized window is protected by haircuts, access limits, and margining rather than by a guarantee of the investment portfolio.

That does not make the structure immune to collective stress. If many participants need cash at once, the settlement buffer may be insufficient, the market may be less executable, and the collateral provider may face a larger volume of advances and margin decisions. Pooling can coordinate the response and reduce duplicated transactions, but it cannot make every asset simultaneously cash-like.