How to Sequence a Multi-Custodian Book When an Advisor Transitions Across Three Custodians
Sequence a multi-custodian transition by opening the receiving custodian accounts in order, then batching repapers by custodian rather than running all three at once. Lead with the custodian holding the largest at-risk AUM or the cleanest account mix, stabilize that flow, then start the next. Sequencing, not parallel chaos, is what keeps SLAs and AUM intact.
An advisor moving a book that sits entirely at one custodian has a hard job with a clear shape. An advisor whose book is split across three custodians, some brokerage at one, advisory at another, a legacy block at a third, has a different problem, and the mistake most teams make is treating it as three transitions running in parallel from day one. Open all three fronts at once and you scatter a finite team across three custodian onboarding processes, three sets of forms, three sets of exceptions, and three status boards, and the thing that slips is the coordination that keeps AUM from leaking during the move.
Multi-custodian books are increasingly common, with a meaningful share of RIAs now using two or more custodians, so this is not an edge case. The skill is sequencing: deciding the order in which custodians get onboarded and repapers get batched so the team works a stabilized flow rather than firefighting three at once. Done well, sequencing turns a three-front scramble into a series of controlled waves that protect the book and the timeline both.
Why a multi-custodian book cannot move as one transfer
Start with the constraint that forces sequencing. Each custodian relationship is its own onboarding: its own new-account paperwork, its own ACATS receiving setup, its own data formats, its own quirks in how it wants forms completed. A transfer is a bilateral thing between a delivering firm and a receiving firm, so a book spread across three custodians is really three related but separate transfer streams, each with its own setup cost and its own exception profile.
That means the up-front work does not parallelize cleanly. Opening receiving accounts at three custodians simultaneously triples the setup load in the exact window when the team is also handling client communication and resignation logistics. And the exceptions do not parallelize either: a NIGO pattern at custodian A is different from the one at custodian B, so a team running all three at once is context-switching between three different reject vocabularies instead of building fluency in one. The core repaper mechanics are the same across custodians, and the foundation for running any of them cleanly is the standard transition discipline in the advisor transition runbook that turns every repaper into a documented repeatable SOP, but the setup and exception handling are custodian-specific enough that doing three in lockstep multiplies the coordination cost rather than sharing it.
The inventory that has to come first
You cannot sequence what you have not measured, so the first step on any multi-custodian book is a complete inventory that separates the book by custodian and profiles each slice. For each custodian, you need the account count, the account-mix complexity, the AUM, and the share of that AUM that is genuinely at risk of not following the advisor. A well-sequenced transition starts with a complete inventory of current accounts, account types, and complex or restricted positions before any paperwork moves, and on a multi-custodian book that inventory has to be cut three ways.
The inventory does two jobs. It tells you the size and difficulty of each custodian slice, which drives the ordering decision below. And it surfaces the complex positions, restricted stock, trusts, alternatives, held-away assets, that carry the most reject risk and the longest timelines, so you know before you start which slice will be slow. Building that profile from statements by hand across three custodians is exactly the kind of volume that makes transcription the bottleneck, which is where AI document intelligence earns its place by extracting positions and account detail from each custodian's statements reliably, so the inventory is accurate rather than a rounded guess that misses the fraction or the restricted lot.
How to decide the order
With the inventory in hand, the ordering decision comes down to a few competing priorities, and the right answer depends on the book. There are two defensible lead strategies, and they optimize for different things.
Lead with the largest at-risk AUM when retention is the dominant concern. The AUM most likely to leak is the AUM you should stabilize first, because every day that at-risk money sits unmoved is a day a competitor can reach the client. Moving the highest-risk slice first shortens its exposure window even though it may be the harder slice to work.
Lead with the cleanest, simplest slice when speed of early wins and team fluency matter more. Starting with the custodian holding the most straightforward account mix lets the team complete a full transfer cycle quickly, build confidence, and establish a rhythm before tackling the complex slice. Early clean completions also give the advisor visible progress to show clients.
| Ordering strategy | Lead with | Best when | Main risk |
|---|---|---|---|
| Retention-first | Largest at-risk AUM slice | High competitive pressure, big accounts exposed | Hardest slice first can slow the start |
| Momentum-first | Cleanest, simplest account mix | Team needs fluency, advisor needs early wins | At-risk AUM waits longer |
| Complexity-isolated | Simple slices batched, complex slice last | One custodian holds most of the hard positions | Longest-timeline slice finishes late |
The wrong strategy is no strategy: starting all three custodians at once, which optimizes for nothing and maximizes coordination load. Whichever lead you pick, the principle is the same. Stabilize one custodian flow before fully opening the next, so the team is deepening fluency in one exception vocabulary at a time rather than juggling three. The inputs that let you predict each slice's timeline and cost before you commit to an order are the same ones a transition model uses, worked through in transition modeling that turns a book's shape into a timeline and an AUM-at-risk number.
Batching the repapers within each custodian slice
Sequencing across custodians is the top layer; within each custodian slice there is a second layer of batching. Once a custodian's receiving accounts are open, group that slice's repapers by similarity so the team works like accounts together: clean individual and joint accounts in one batch, retirement accounts in another, trusts and complex registrations in another. Batching by account type lets a specialist build a rhythm on one form pattern instead of switching cognitive context on every account, which lowers the error rate that feeds NIGO rejects.
This is where multi-custodian sequencing connects to concurrency management, because within a stabilized custodian flow you are running many repapers at once and must avoid the collisions that come with volume, including duplicate submissions and status confusion. The discipline for running many overlapping repapers without missing turnaround commitments is the same regardless of which custodian slice you are working, and it is detailed in running concurrent advisor transitions and how a recruiting desk sequences overlapping repapers without missing SLAs. The point of sequencing across custodians is precisely to keep this within-slice concurrency manageable: one custodian's worth of concurrent repapers is a controlled load, three custodians' worth started simultaneously is not.
Client communication has to track the sequence
The sequence is not only an operational plan; it sets the client-communication schedule, and getting that alignment right is what protects AUM through the move. Clients whose accounts are in the first wave need to hear from the advisor before their accounts move, and clients in a later wave should not be left wondering why their neighbor's account moved and theirs has not. A sequence that the client communication does not track produces exactly the anxiety that lets a competitor's call land, so the communication plan has to mirror the custodian ordering: this custodian's clients hear from us this week, that custodian's clients next.
Nearly a third of RIAs run more than one custodian, and industry guidance on multi-custodian operations, such as Kitces on choosing and working with RIA custodians, underscores that each custodian relationship carries its own operational profile, which is another reason to stabilize one flow and its client communications before opening the next. The way client communication itself preserves AUM through a transition, wave by wave, is a discipline in its own right, and the standard for keeping the book you move is set out in AUM retention and how a client communication sequence holds assets from resignation day to full repaper. The rule-governed transfer framework underneath all of it, common to every custodian, is documented in FINRA's customer account transfer task force guidance.
Sequencing is the whole job on a multi-custodian book
The temptation on a three-custodian book is to prove urgency by starting everything at once, and it is exactly backward. Parallel-from-day-one scatters a finite team across three onboarding processes and three exception vocabularies, and the coordination that protects AUM is the first thing to fray. Sequencing, inventory the book by custodian, pick a lead slice by at-risk AUM or clean mix, stabilize that flow and its client communication, then open the next, turns the same work into controlled waves the team can actually execute. The firms and transition consultants who move complex multi-custodian books without leaking AUM are not working faster on three fronts; they are working one stabilized front at a time in a deliberate order. Running that sequenced approach on a purpose-built advisor transition platform, with one status view across all three custodian streams, is what keeps a three-custodian move from becoming three competing scrambles, and the compounding result of that discipline is the kind of outcome behind the advisor transition case study.
Move a multi-custodian book as a planned sequence of waves, not as three simultaneous transitions, and the same team that would drown running everything at once protects both the timeline and the AUM.
Frequently asked questions
Should I move all three custodians at the same time to go faster? No. Starting all three at once scatters a finite team across three separate onboarding processes, three sets of forms, and three different exception vocabularies, which multiplies coordination cost rather than sharing it. The coordination that keeps AUM from leaking during the move is the first thing to fray under that load. Sequencing, stabilizing one custodian flow before fully opening the next, lets the team build fluency in one exception pattern at a time and generally protects both the timeline and the book better than parallel chaos.
Which custodian should I onboard first? It depends on the book. Lead with the largest at-risk AUM slice when competitive pressure is high and big accounts are exposed, because every day at-risk money sits unmoved is a day a competitor can reach the client. Lead with the cleanest, simplest account mix when the team needs to build fluency and the advisor needs early visible wins. The one wrong choice is no order at all, starting everything simultaneously, which optimizes for nothing and maximizes coordination load.
What has to happen before I can sequence? A complete inventory that separates the book by custodian and profiles each slice: account count, account-mix complexity, AUM, and the share of AUM genuinely at risk of not following the advisor. That inventory drives the ordering decision and surfaces the complex positions, restricted stock, trusts, alternatives, and held-away assets, that carry the most reject risk and the longest timelines, so you know before you start which slice will be slow. Building it accurately across three custodians is where reliable statement extraction matters most.
How do I batch repapers within one custodian? Once a custodian's receiving accounts are open, group that slice's repapers by similarity so specialists work like accounts together: clean individual and joint accounts in one batch, retirement accounts in another, trusts and complex registrations in another. Batching by account type lets a specialist build a rhythm on one form pattern instead of switching context on every account, which lowers the error rate that produces NIGO rejects. This within-slice batching is manageable precisely because sequencing kept you from running three custodians' worth of concurrency at once.
How should client communication line up with the sequence? It has to mirror the custodian ordering exactly. Clients whose accounts are in the first wave hear from the advisor before their accounts move, and clients in a later wave should be told where they are in the plan so they do not wonder why a neighbor's account moved and theirs has not. A sequence the communication does not track produces the anxiety that lets a competitor's call land, so the plan is this custodian's clients this week, that custodian's clients next, wave by wave, until the whole book has moved.
