Asset Management Strategies Case Studies
Finance5 Case StudiesUpdated Sep 2026
Asset management is mostly a coordination problem. Portfolio teams, fund operations, client service and compliance touch the same positions and deadlines, so a slip in one place reaches the rest. The five cases are set in a fictional group of firms. Each starts with a symptom that looks like a people or system issue and proves to be both. In every case we examine roles, service levels, workflow data, stakeholder expectations and delivery targets, because those tend to fail together. The aim is to separate symptoms from causes and choose changes that hold past the first month.
Case Study 1
Fund administration team misses monthly reporting targets
Ashcombe Fiduciary Services, a fund administration unit that calculates net asset values and produces investor reports for about 140 funds across three fund families.
The situation
The fund accounting team at Ashcombe calculates daily net asset values, processes corporate actions and assembles monthly investor packs for about 140 funds. Over two consecutive quarters, the share of packs delivered by the agreed day fell from 91 percent to 74 percent. Nobody had been formally assigned the pricing exceptions queue, so analysts picked those items up between other tasks. Portfolio managers were sending trade files after cutoff, and fund controllers kept routing urgent reconciliation requests straight to individual analysts. The team had three working definitions of delivered: the internal checklist, the client service agreements and the controllers' sign-off calendar. Each was reasonable on its own.
The challenge
Hanna, the team lead, had been asked by the head of fund operations to explain the slippage and bring a plan to the next steering meeting. The team had strong individual performers but no shared picture of where time went. Some analysts blamed the portfolio managers, others blamed the queue, and Hanna suspected both were partly right. Headcount was frozen for the year, so the plan had to work with the people already on the team.
Analysis
Teams grounded in Asset Management Strategies usually begin with the work queues rather than the people working them, because accountability only holds when each queue has one named owner and a service level everyone can see. Hanna pulled three months of ticket timestamps and found that the pricing exceptions queue had no owner for roughly a third of its items. Those items sat for about four working days before anyone touched them, while urgent reconciliation requests closed within the day. The difference was not effort. Reconciliation requests had a named requester and a visible deadline, and exceptions had neither.
The stakeholder picture was just as tangled. The client agreements counted a pack as delivered when it reached the investor by the second business day, the internal checklist marked it done after review, and the controllers measured it by their own sign-off. Each definition produced a different percentage from the same month's data. Before changing anyone's workload, Hanna asked the three groups to agree on one target per pack type and one escalation path for late inputs. The most useful finding was that portfolio managers were late mainly on trade files submitted after cutoff. That pointed to a cutoff rule, not a performance problem on the team.
- First working weekOwners named for every queue; delivery definitions drafted with controllers.
- Second monthTrade file cutoff and same-day exception form go live.
- End of second monthPacks back in the mid-eighties; exception age under one day.
- End of the quarterWeekly queue review runs without the operations lead present.
Actions taken
- Assign a single named owner to each work queue, including the pricing exceptions queue, and publish the owner list on the team board within the first week.
- Agree one written definition of delivered for each pack type with the client service leads and the fund controllers, then measure every month against that definition only.
- Set a trade file cutoff with portfolio managers and route late submissions through a same-day exception form that records the requester so delays become visible to the people causing them.
- Review the weekly queue data in a fifteen-minute standing meeting, with owners reporting item age and reopen counts, and escalate any item older than two working days to the operations lead.
Results
Within two months the share of packs delivered by the agreed day returned to the mid-eighties, and the average age of exceptions fell from about four working days to under one. The cutoff rule did most of the work, cutting late trade files by roughly half. Hanna did not add headcount. Two analysts moved routine reconciliation requests onto a shared intake form, which freed a few hours each week. The team still misses occasionally, but each miss now appears on the board with an owner and a reason, which is what the steering group asked for.
Questions for discussion
- Which of the three delivery definitions would you have chosen as the single measure, and what would that choice have cost the other two groups?
- What would you check in the second month to confirm the named-owner model is working, beyond the on-time percentage?
- A senior controller keeps sending urgent work straight to one analyst. What would you do before raising it in a steering meeting?
Case Study 2
Pension client escalation over late attribution reporting
Pellham Street Pension Advisers, a manager of liability-driven fixed income portfolios for three municipal pension plans.
The situation
The plan's chief financial officer, Dana, wrote twice in one quarter about the attribution report that Pellham Street produces for the plan's board. The report was due by the fifth business day and arrived on the eleventh. When it did arrive, two bond positions carried par values that did not match the custodian statement. Dana then escalated to the managing director, saying the firm's reporting had cost the board a meeting. Internally, the portfolio manager had been reconciling the holdings herself, while the client service team forwarded Dana's emails to her without logging them. Dana wanted a call within the week, a corrected file and a date she could give her board.
The challenge
Teo, the client relationship director, had three days before the board packet closed. He needed to decide what to promise, which commitments the reconciliation backlog could support, and how to repair the relationship without making operations look careless. He also suspected that part of Dana's frustration came from an earlier fee discussion that he had not been part of and that the file did not explain.
Analysis
The first problem was structural. The service agreement set a delivery date for the attribution report but named no owner for the holdings reconciliation that fed it, so a custodian mismatch had no clear route to resolution. The portfolio manager was doing the work informally, the client team logged nothing, and Dana could only see status by asking again. Teo started by writing down what the agreement actually promised: the report date, the accuracy standard, the response time for questions, and whether any credit applied to late delivery. The report date, accuracy standard and credit terms were clear. Response time was not defined anywhere, which is why Dana's emails felt ignored even when they were answered within the week.
Dana's needs had changed since the last review. The board now used the attribution report to decide on rebalancing, so a late file carried a cost it had not carried before. Teo separated what could be fixed in three days from what needed a longer commitment. A corrected holdings file and a written explanation of the par value difference were achievable before the packet closed. A permanent fix to reconciliation, with weekly status notes, was something the firm could measure. He held back on a return to the original schedule until two clean cycles existed. Dana valued commitments she could check more than reassurance.
Actions taken
- Call Dana the same day, confirm in writing what the firm will deliver before the packet closes, and ask which board date matters most so commitments are sized to it.
- Have the portfolio manager and a reconciliation analyst deliver a corrected holdings file with a one-page note on the par value difference, reviewed by a second person before release.
- Name one service owner for the plan account with a direct line, and set a response standard of one business day for acknowledgment and five for substantive answers.
- Send a short status note every Friday for the next quarter covering report timeliness, open reconciliation items and any variance above the agreed threshold.
Results
The corrected file reached Dana two days before the packet closed, and the board saw the explanation with the report. Dana then asked for the same status note format on the plan's other accounts, the clearest sign trust had returned. Over the next quarter the report arrived on the sixth business day on average and every Friday note went out on time. One reconciliation item stayed open longer than planned, and Teo reported it in the note before Dana had to ask. The fee question came up at the next review and was handled against a written record.
Questions for discussion
- Where would you draw the line between a commitment the firm can keep under the current backlog and one that only sounds reassuring?
- How would you handle a client's sense of a past grievance when the file contains no record of it?
- If Dana asked for a different owner on the account, what would you need to know before agreeing?
Case Study 3
Trade break loop between operations and custodian
Tallow Marsh Fund Administration, which settles trades and reconciles cash for a group of pooled funds.
The situation
Tallow Marsh settles about 6,400 trades a month for a group of pooled funds. Roughly one trade in twenty-five ends up in a break, meaning the custodian's record and the internal record disagree on price, quantity or allocation. Each break is found by an operations analyst, returned to the front office by email, corrected by hand and resubmitted, and the same trade often breaks again in the next cycle. Average break age stood at three and a half days, and about one break in five involved a trade that had already been corrected once. Month-end was the worst point, when controllers needed clean positions before the valuation run.
The challenge
Kofi, the operations manager, had been asked to halve break age within the quarter without new system spend. The team was stretched and opinions diverged on the root cause. The front office thought the custodian's file format was the problem. The custodian pointed at late bookings. Kofi had two options on the table, a change to the allocation field mapping and a daily pre-match with the front office, and no evidence yet on which would last.
Analysis
Kofi mapped the process over two weeks by tagging every break with the step where it first appeared. The map had eleven handoffs, and four of them were email or spreadsheet transfers with no timestamp. The break return was the longest single handoff, because a break could sit in an analyst's inbox until the front office trader was free to answer. Tagging the breaks showed that about 60 percent originated at matching rather than booking, which contradicted the front office view. Within matching, the largest share came from one allocation field that the custodian read differently from the internal system.
The three groups wanted different things from the same process. The front office wanted trades booked on time, the custodian wanted clean files by its cutoff, and the fund controllers wanted no open breaks at valuation. Kofi compared the two options against those expectations. Changing the allocation mapping was cheap and fixed the largest reason code, but it depended on the custodian's schedule. A daily pre-match started quickly and handled late bookings, but took about twenty minutes of front office time each day. He chose to start with the mapping on one fund and keep the pre-match in reserve for large trades.
- Booking Front office books trades; late bookings create the first mismatches.
- Matching Overnight custodian match flags mapping mismatches as ordinary breaks.
- Break return Breaks emailed to analysts, who key corrections by hand.
- Rework Corrected trades resubmitted, often re-breaking in the next cycle.
Actions taken
- Stop the email return loop by logging every break in one shared tracker with a single owner, a first-response target and a reason code, starting in the first week.
- Fix the allocation field mapping with the custodian first, testing on one fund before extending it, since that reason code accounts for most repeat breaks.
- Hold a fixed ten o'clock pre-match with a front office representative for trades above an agreed size, with a written list of open items that both sides initial.
- Track break age and rework count weekly against the baseline for one quarter, and defer any automation decision until the remaining breaks have been measured.
Results
Break rate fell from about one trade in twenty-five to roughly one in forty within the quarter. Average break age dropped to just over one day, and rework on repeat breaks roughly halved. The mapping fix drove most of the gain on the pilot fund and then went to the rest of the group. The pre-match stayed in reserve for large trades, which kept front office time to a few minutes a day. Kofi's team reports age and rework to the controllers weekly, and automation stays open until the remaining breaks are measured.
Questions for discussion
- How would you decide whether the pilot fund's improvement came from the mapping fix or from the attention the project received?
- What would make you move the pre-match from reserve back into the daily routine?
- Which handoff in this map would you remove first, and what would you need to confirm before doing so?
Case Study 4
Liquidity policy rollout across six open-ended funds
Kestrel Point Investment Management, a manager of six open-ended funds with a small risk and compliance team of its own.
The situation
A revised liquidity risk policy at Kestrel Point requires every open-ended fund to bucket its holdings daily by estimated time to liquidate and to run a monthly stress test on the results. The risk system that produces the buckets is being upgraded at the same time. Portfolio managers were skeptical that the buckets reflected how their markets actually traded. Analysts were worried about the extra forty minutes per fund that the pilot had added. Compliance wanted evidence that the policy was being used, not just adopted. The 9:30 risk report, which the investment committee reads before the open, still had to go out on time.
The challenge
Priya, head of risk, had to get uptake before the quarter closed and show it in a form the risk committee would accept. The firm's previous liquidity policy had been formally adopted but was rarely used in practice, and Priya wanted to avoid repeating that. Her sponsors expected the policy live on schedule, the portfolio teams wanted flexibility in how they built buckets, and compliance wanted measurable proof of use.
Analysis
The objections were specific once they were sorted by team. Portfolio managers objected to the liquidity assumptions for their own markets, and their concerns were about model inputs rather than the policy itself. Analysts objected to the time cost, which the pilot confirmed. Compliance objected to the absence of evidence. Priya pulled the pilot's usage logs and found that three of the six funds had completed the bucketing screen by 9:30 on fewer than half the days, usually because the upgraded system needed a manual refresh that nobody owned. That turned a dispute about policy into a question about a single step, which was easier to fix.
Priya sequenced the steps so each one gave the next something to build on. The first step was an owner for the refresh and a written response to each team's objections within two weeks. Next came early adopters: one champion per team, given protected time to build a template. The wider rollout followed only after a month of parallel running, with a daily comparison that showed where the old and new methods disagreed and why. Uptake was measured at fund level: completion by 9:30, count of manual overrides, and stress test exceptions. Those three numbers went to the risk committee each month, so adoption was judged on use.
Actions taken
- Hold a closed session with each portfolio team to record its objections in writing, and publish a response to each one within two weeks, including the ones the policy will not change.
- Name one champion per team from the early-adopter group and give each two weeks of protected time to build the template and answer questions before the wider rollout.
- Run the new bucketing in parallel with the old process for one month, with a daily comparison report that shows where the two methods disagree and why.
- Set uptake measures at fund level, covering on-time completion by 9:30, manual overrides and stress test exceptions, and review them in the risk committee each month.
Results
Five of the six funds completed daily bucketing by 9:30 on at least nine days in ten by the end of the quarter, up from roughly half in the first week. Manual overrides fell by about two thirds once the template was shared. The parallel run exposed one real problem, an overly generous liquidity assumption for a small credit holding, which the portfolio manager revised once the comparison showed the gap. The sixth fund stayed a holdout, and Priya reported its specific objection to the committee. Compliance now has three months of usage data to review.
Questions for discussion
- Which of the stakeholder groups in this matrix would you expect to move first, and what evidence would you use to test that expectation?
- How would you respond if a champion began to resist the policy after the template was built?
- What usage pattern would convince you that the parallel run could end early without weakening the rollout?
Case Study 5
Choosing two initiatives from four competing proposals
Northmere Endowment Advisory, which manages investment portfolios and model services for a set of university and foundation clients.
The situation
Northmere's investment committee has budget for two initiatives in the coming year and four proposals on the table. The first is a private credit sleeve for advisory clients. The second is a rebuild of the performance attribution system, which the operations lead says is holding back quarterly reporting. The third is an ESG data feed that the research team wants for screening. The fourth is an expansion of the model portfolio service, which two pilot advisers have used for a few months. Each budget range is wide, and client demand data is partial for all four.
The challenge
Sana, the chief investment officer, wanted the decision made in the next committee cycle with the reasoning written down. The chief operating officer, Marcus, wanted the attribution rebuild, and the head of client service, Ines, favored the model portfolios. Sana leaned toward private credit. Nobody had agreed on how to compare proposals whose benefits showed up in different places. The committee needed criteria everyone accepted, a view of risk for each option, and a way to commit money without locking in the rest of the year.
Analysis
The committee began by agreeing on five criteria, weighted before any proposal was rescored: client impact, delivery risk, cost range, reversibility and data quality. The weights went into the decision memo before any results arrived. Each proposal was then scored from one to ten on each criterion, with a confidence rating attached. The attribution rebuild scored well on delivery risk since the vendor was known, but its cost range was wide. Private credit scored high on client impact in the sponsor's view, yet the confidence rating was low because demand came from a handful of advisers. Scoring confidence alongside the score made the gaps visible.
The stakeholder expectations were what made a staged decision necessary. The operations lead wanted a dated plan, the client team wanted proof of demand, and Sana wanted a route to private credit. A staged decision gave each of them a measurable target. The committee funded the attribution rebuild in full, with a fixed scope and a delivery date, and funded a capped six-week pilot of the model portfolios. The private credit proposal went into a documented queue with a re-review date and a small discovery budget, released only if the stated assumption about adviser demand could be tested.
Actions taken
- Agree five weighted criteria with the four sponsors before any proposal is rescored, and record the weights in the decision memo so they cannot shift after results arrive.
- Attach a confidence rating to every estimate, and require each sponsor to name the single assumption that would change the score if it proved wrong.
- Fund the attribution rebuild with a fixed scope, and run a capped six-week pilot of model portfolios with a stage gate at the end of the second quarter tied to adoption and error counts.
- Hold the private credit and ESG proposals in a documented queue with a re-review date, and release discovery budget only when the stated assumption has been tested.
Results
The committee approved the attribution rebuild and the six-week model pilot. The rebuild came in near the upper end of its cost range and delivered the dated plan the operations lead had asked for. The pilot had three advisers live and passed its stage gate on adoption and error counts. The private credit proposal returned to the queue with a client survey showing demand from only a small share of advisers, which settled the question for the year. Nobody claimed certainty, but each choice had a written reason and a checkpoint.
Questions for discussion
- Which criterion would you weight most heavily for a firm whose clients are asking for private credit but whose demand data is still thin, and how would you defend that weight?
- What would make you reopen a queued proposal, and how would you keep that trigger from becoming a way to avoid a decision?
- If the model pilot had failed its stage gate, what would the committee owe the two advisers who had already adopted it?
Learning Resources
Finance CASE STUDIES
China, India, United States, Indonesia, Pakistan, Nigeria, Brazil, Bangladesh, Russia, Japan, Mexico, Philippines, Vietnam, Egypt, Turkey, Germany, United Kingdom, France, Italy, Canada, South Korea, Australia, Spain.