Who Knows What in Debt and Investment Management, and Where It Goes Wrong

The dealer knows why a trade was done. Settlements knows what actually moved. Accounting knows how it must be recorded, and risk knows which limits apply. Trouble starts at each handoff, where one role assumes another already holds context it was never given.

The dealer holds the intent

A dealer placing a deposit, rolling a facility or buying a forward knows the reason behind it. Perhaps the trade hedges a specific supplier payment. Perhaps it parks surplus cash until a bond coupon falls due. Maybe it was a favour to a bank that has been generous with credit lines.

Very little of that reason reaches the trade ticket. The ticket shows counterparty, amount, rate and dates. The purpose lives in the dealer's head or in a chat thread with the bank.

This matters later. When accounting needs to designate a hedge, it needs the link between the derivative and the exposure it covers. If that link was never written down at the time of the trade, someone reconstructs it after the fact. Auditors notice reconstructed documentation.

Settlements sees what the bank sees

The back office confirms trades with counterparties and matches them to what arrives on bank statements. It is the first function to learn that a confirmation came back with a different value date, or that a bank applied the wrong standing instructions.

Settlements rarely knows whether a mismatch is a typo or a real dispute. It also tends not to know which banking relationships are sensitive. A firm query sent to a bank the dealer is courting for a new facility can cause friction nobody upstream expected.

The reverse gap is just as common. Dealers often do not hear about failed or late settlements unless the amount is large. Patterns of small breaks with one intermediary go unseen at the desk for a long time.

Accounting knows the rules but not the story

Accounting knows how fair value movements flow through the books and which classification each instrument needs. It knows what the auditors asked about last time. What it lacks is a view of trades in flight. An interest rate swap agreed late in a period may not appear in the system until after close.

The transaction reports that accounting produces are built from booked data. If the treasury system and the general ledger disagree, accounting usually discovers it during reconciliation, long after the dealer has moved on. Investment classification is another sore point. A dealer may treat a holding as available to sell at any moment. Accounting may have recorded it as held to maturity, which carries consequences if it is sold early.

Risk and policy owners know the limits

Whoever owns the investment policy knows the approved instruments, the credit ratings required and the concentration ceilings for each issuer. Risk monitors those limits.

The desk often knows the policy only in outline. It may not know that a counterparty was downgraded last week, or that a parent group and its subsidiary count as one issuer for exposure purposes. Breaches of issuer limits usually come from this kind of gap. They are rarely deliberate.

Policy owners, for their part, can be far from the market. A policy written when certain money market funds were plentiful may now leave the desk with few eligible places to put cash. Dealers work around it quietly instead of asking for a change.

The business knows when cash will really move

Liquidity management depends on forecasts from operating units. Those units know about the delayed customer payment, the capital project that slipped and the acquisition being negotiated in confidence. Treasury sees a submitted number.

Forecasters seldom understand what treasury does with their figures. They do not see that an optimistic receipt estimate leads to a longer investment tenor, which then has to be broken at a cost when the cash fails to arrive. Without that feedback, forecast quality never improves.

Foreign currency is a special case. A subsidiary may agree a contract in a foreign currency without telling anyone centrally. Treasury learns of the exposure when the payment request lands, too late to hedge it calmly.

Questions to ask the people who run it

The documented process usually describes how tickets flow. It says little about the conversations that make the process work. Ask:

• When a trade is done to hedge something, where is that link recorded, and who could find it without asking the dealer?
• What happens when a confirmation does not match? Who decides whether to escalate to the bank?
• Which trades regularly arrive in the ledger after period close, and why?
• How does the desk learn that an issuer has been downgraded or that a limit is close to full?
• Are there instruments or counterparties the desk avoids because the policy is unclear, or uses because nobody has said no?
• Do the people who submit cash forecasts ever see how accurate they were?
• Which bank relationships carry history that is not written anywhere?
• If the most experienced dealer left tomorrow, which knowledge would leave too?

The answers tend to show that the process runs on a handful of people who quietly carry context between functions. Any redesign that removes those informal channels without replacing them will produce errors that the documented controls were never built to catch.

APQC's Process Classification Framework® (PCF) is an open standard developed by APQC, a nonprofit that promotes benchmarking and best practices worldwide. To download the full PCF or to view definitions and measures, please visit www.apqc.org/pcf.