Merrill Lynch had the largest net loss of any firm family: 895 investment adviser representatives

1. Merrill Lynch had the largest net loss of any firm family

Competitive moves, October 21, 2025 to October 20, 2026, compared with October 21, 2024 to October 20, 2025. Data through October 20, 2026.

1,417 IARs left Merrill Lynch for another firm family and 522 joined, a net loss of 895. Of those who left, 41% went to three firms: J.P. Morgan (251), Wells Fargo Advisors (185) and Morgan Stanley (149). Merrill Lynch to J.P. Morgan was the busiest firm-to-firm route in the market.

Departures were about flat (−1% to +3%) on the year before: 1,381 in the prior 12 months, rebuilt from the same records.

2. Commonwealth's departures equaled 17.9% of its IARs at the start

Competitive moves, October 21, 2025 to October 20, 2026. Data through October 20, 2026.

Commonwealth Financial Network had 3,139 IARs on October 21, 2025. The window recorded 561 competitive departures, equal to 17.9% of that starting count, or more than one in six. That is the highest departure rate of any firm family with 1,000 or more IARs at the start; the lowest in that group was State Farm's, 0.3% (20 of 6,164). The most common destinations were Raymond James (92), Cetera (76) and Kestra (74).

The records show when departures rose, not why. Before LPL Financial announced its acquisition of Commonwealth on March 31, 2025, Commonwealth recorded 23 to 41 competitive departures a quarter. Departures peaked at 259 in Q3 2025 and were 119 in Q3 2026, a figure that late filings can still raise. The bars are calendar quarters, so they do not add up to the window's count.

223 of the 561 departures (40%) were in registration clusters: three or more advisors who left the same firm for the same destination firm, in the same state, each arriving within three days of the one before. Shuffling the same people's arrival dates puts about 11 departures (2%) in clusters by chance. A cluster shows registration timing. It does not show that the people worked together or decided together.

The 17.9% covers the study window only. It is not a count since the announcement: departures from March 31, 2025 to October 20, 2025, including the peak quarter, fall before the window.

3. RIA firms on Continuum's list gained more than a year earlier; wirehouses and insurance-affiliated firms lost the most

Competitive moves, October 21, 2025 to October 20, 2026, compared with October 21, 2024 to October 20, 2025. Data through October 20, 2026.

This view covers competitive moves with a classified operating model at both ends, 49.4% of all competitive moves. RIA firms in Continuum's classification list had a net gain of 604, against 473 in the prior 12 months (489 after the rebuild allowance described under Method). Wirehouses had the largest net loss (−748), followed by insurance-affiliated firms (−598). Independent broker-dealers had the largest net gain (+826). The wirehouse group here includes J.P. Morgan.

The classification list covers 164 firms, 31 of them RIAs. An RIA that is not on the list is unclassified, so this view leaves out most moves to independent RIAs. See the Advisor Channel Flow report for the full matrix.

4. J.P. Morgan and Wells Fargo Advisors recorded more arrivals than a year earlier

Competitive moves, October 21, 2025 to October 20, 2026, compared with October 21, 2024 to October 20, 2025. Data through October 20, 2026.

J.P. Morgan recorded 971 arrivals, against 723 in the prior 12 months: about +22% to +34%. Wells Fargo Advisors recorded 986, against 747: about +24% to +32%. Merrill Lynch was the largest single source of arrivals at both.

5. Among the largest firm families, departure rates ran from 3.1% to 5.9%

Competitive moves, October 21, 2025 to October 20, 2026. Data through October 20, 2026.

Of the 10 firm families with 10,000 or more IARs at the start of the window, the lowest departure rate was Edward Jones, 3.1% (717 of 22,776) and the highest was Ameriprise, 5.9% (717 of 12,182). A departure rate is competitive departures divided by IARs at the start of the window. It shows who left, not why, and it is not a measure of how well a firm keeps its advisors.

6. Across the market, competitive moves rose about 5% to 8%

Competitive moves, October 21, 2025 to October 20, 2026, compared with October 21, 2024 to October 20, 2025. Data through October 20, 2026.

SEC records show 22,816 competitive moves in the window, against 21,117 in the prior 12 months on the same definition: about +5% to +8%. Both windows leave out confirmed corporate events: 1,052 moves in this window and 2,757 in the prior one, where Prudential Advisors' move onto LPL Financial's platform accounts for 2,128. Counting those events, the two windows were about level: 23,868 moves against 23,874. The rise appears only once corporate events are left out of both windows.

Across the 116 firms with enough departures to test, 16% of departures were in registration clusters, against 3% from chance timing. Clusters show registration timing only.

How this differs from other counts

Other published counts of advisor movement count different people, with different rules, over different periods, so their totals and rates differ from these.

• Who is counted. This study counts investment adviser representatives in SEC adviser records. Counts built on FINRA broker registrations, or on every licensed representative, also include broker-only and insurance-licensed people, so they are usually larger. Counts limited to experienced advisors, or to moves announced in the trade press, are usually smaller.
• What counts as a move. A move here is a change to a different firm family, counted once per person. Moves inside a family are left out. So is a change of registration type that leaves a person's adviser firm unchanged, such as dropping a broker-dealer registration. Counts that treat that change as a channel move show much larger RIA gains.
• Corporate events. Confirmed acquisitions, platform conversions and internal mergers are left out, such as Prudential Advisors' move onto LPL Financial's platform. Counts that include them show very different totals for the firms involved.
• Firm families. A family counts every IAR registered with its firms, whatever their role or office. Merrill Lynch's count, for example, is every IAR registered with Merrill Lynch, Pierce, Fenner & Smith. A firm's own figures often cover one business line, or only experienced producing advisors.
• Net flow is not headcount. Net flow counts competitive moves in and out. A firm's headcount also changes with new entrants, retirements, people leaving the industry and corporate events, so its reported change can differ in size and even in sign.
• Period. The window is the 12 months from October 21, 2025 to October 20, 2026, not a calendar year or a half-year.
• Commonwealth. The Commonwealth rate covers the study window only. Departures between LPL Financial's announcement on March 31, 2025 and October 20, 2025 fall before it, so counts that start at the announcement are higher. The rate is not the reverse of a retention rate: it counts only moves to other firm families, and its base is every IAR registered with Commonwealth. Retention stated as a share of assets measures something else. At the end of the window most of Commonwealth's IARs were still registered with Commonwealth, so a later move onto LPL Financial's own platform is not in these figures.

Method

The source is the SEC's Investment Adviser Public Disclosure (IAPD) feed of October 21, 2026, processed with the same code that builds Continuum's Market Intelligence pages. The methodology page explains each rule in more depth.

Competitive move

A registration that ends at one firm and begins at a firm in a different firm family. Moves inside a family, such as rebrands and internal transfers, are left out, and so are registrations that resume after a break of more than 180 days.

Confirmed corporate events

Acquisitions, platform conversions and internal mergers in which whole groups re-register at once. The study excludes 22 confirmed events, each supported by a public source. Their transfer days are left out of both windows.

One count per person

Each person counts once in a window, at their latest move.

Window

Arrivals from October 21, 2025 to October 20, 2026. The prior year is October 21, 2024 to October 20, 2025.

IARs at the start

Investment adviser representatives registered with a firm family when the window opened on October 21, 2025, rebuilt from registration start and end dates.

Departure rate

Competitive departures in the window divided by IARs at the start. The Firm Churn report follows the starting roster instead, so its rates differ slightly.

Net flow

Arrivals minus departures.

Operating model

The channel of each firm in Continuum's reviewed classification list. A move counts in this view only when both firms are on the list.

Registration cluster

Three or more advisors who left the same firm for the same destination firm, in the same state, each arriving within three days of the one before. A firm's share is published only when it had 20 or more departures and the share is at least twice what chance timing produces. Chance is measured with 1,000 shuffles of the same people's arrival dates, and a published share has less than a 5% chance of arising that way. Otherwise the share is reported as not distinguishable from chance timing.

Year-over-year ranges

The prior year is rebuilt from today's records, which have lost the people who left the industry since. To size that loss, the same prior-year counts were rebuilt from the records of June 17, 2026, 126 days earlier: the market total fell 1.2%, and the firm counts compared here fell up to 3.6%. Each range runs from the change as rebuilt today to the change after raising the prior-year count by that measured fall (at least the market-wide fall), carried over the full year between the windows. A change inside 5% either way at both ends is called about flat.

The queries, in plain language

1. Read every individual record in the SEC feed of October 21, 2026, with each person's history of firms.
2. For each person, take the latest move that arrived on or before the end of the window, and keep it if it arrived inside the window.
3. Leave out moves within a firm family, moves after a long break and moves on a confirmed corporate event's transfer days.
4. Count departures by the family left and arrivals by the family joined. Net flow is arrivals minus departures.
5. Count each family's IARs on October 21, 2025 and divide departures by that count.
6. Repeat the second to fourth steps for the prior year, from the same records and from the records of June 17, 2026.
7. Find registration clusters and test each firm's share against shuffled arrival dates.

Filing lag

Registrations are sometimes filed weeks or months late, so recent counts can still rise. Treat the latest quarter as provisional.

Limitations

• IARs are not the same as producing advisors. A firm's IAR count includes supervisory, home-office and administrative staff.
• IAPD covers IARs only. Broker-only registered representatives, who hold a FINRA registration but no adviser registration, are not in these records.
• People who left the industry drop out of the records together with their history. Earlier periods are understated, and moves out of the industry are not counted.
• The records show where and when people re-registered. They do not show why.
• A registration cluster shows timing. It does not show that people worked together or decided together.
• The latest quarter is provisional while late filings arrive.

Data

Download the firm-level table (CSV). It has one row per firm family in Continuum's classification list: operating model, IARs at the start, departures, arrivals, net flow, departure rate and registration-cluster share. Where no share is shown, the row says why. The findings file (JSON) holds every figure on this page with its definition and window.

Cite this study

Continuum Search Group, Advisor movement in the 12 months to October 2026, data through October 20, 2026, https://www.continuumsearchgroup.com/market-intelligence/study-advisor-movement-october-2026.

For press

Press contact: Jack Kerstiens, Co-Founder, Continuum Search Group, jack@continuumsearchgroup.com. Every figure on this page names its window and data date, and the CSV holds the firm-level counts behind it.