An investment proposal generator that reads every held-away statement and cites every fee, for the advisor to approve
A cited proposal in minutes, with the cost of every holding and every alternative on file.
See one case, screen by screen ↓An investment proposal generator, also called investment proposal software or an investment proposal tool, turns a prospect's statements into a proposal an advisor can present: today's holdings and costs, a recommended portfolio and the comparison between them. Investment Proposal Generator does it with seven AI agents that read held-away statements from any custodian, price every holding from its own prospectus, X-ray overlap and concentration, test risk fit and draft a cited fee and risk proposal with every alternative on file. The advisor approves and a registered principal reviews before the client sees it.
AI in asset management: a buyer's guide to agentic AI for investment managers →
Updated
An investment proposal generator is only as good as the statements it can read. A prospect brings held-away assets from other firms: an advisory wrap account, a rollover IRA, sometimes an inherited account at a private bank abroad, written in another language. Today an advisor or an associate retypes every position into the firm's proposal generation software, looks up fees one fund at a time, guesses at overlap between large-cap funds and writes the narrative in a word processor, often the night before the meeting.
The costs that matter hide where nobody has time to look: a C share held long enough to convert, a 1.25% wrap fee outside the funds, a custody fee on page five of a German statement. And the file has to stand up afterwards. Reg BI's care obligation expects the costs and reasonably available alternatives behind each recommendation, the account type included; a rollover recommendation made under PTE 2020-02 needs its documented reasons; a hypothetical savings line needs its disclosure and assumptions under the SEC Marketing Rule; and the firm's procedures call for a registered principal to review it before the client sees a page.
Estimated split for one prospect with three statements, by hand and with the solution, in hours.
Seven agents read, price, X-ray, test and draft every proposal; the advisor approves and a registered principal reviews.
Lorraine Vance, 61, a retiring anesthesiologist in Scarsdale, New York, brings Quillfield Private Wealth three statements worth $2,371,640: a Harborline Securities wrap account, a Crestway rollover IRA and her late husband's account at a Zürich private bank, in German. She meets her advisor, Ethan Marsh, on Thursday, October 22, at 2:00 p.m. It is Tuesday morning. Here is what happens next, screen by screen, in the working solution.
Ethan's home screen shows what the agents did while he slept: the Statement Reader read Lorraine's three statements, the Cost Lookup agent priced her at 1.55% all-in, $36,776 a year, the Proposal Writer drafted v3 and the Marketing Review agent flagged three lines. Each message cites its source and opens it. Under “Needs you” there is one clear next step for each of three prospects.
“Draft proposal ready. One line needs its hypothetical disclosure before you can approve.”
Three statements, 29 pages, 412 values read, and 3 of 3 totals reconciled to the statement before anything else runs. The Depotauszug from Privatbank Aldrich & Söhne AG is shown in German beside the English, and CHF 414,407.05 is converted at the September 30 closing rate of 1.2480: $517,180. On page 5 it finds the Depotgebühr, a custody fee of 0.35% a year, CHF 1,450.42 ($1,810), charged outside the funds.
“This custody fee is charged by the bank, outside the funds. It is not in the fund expense ratios, so the Cost Lookup agent adds it to her all-in cost.”
Values below the 95% confidence threshold stop here; nothing is guessed. Page 7 of the Harborline statement was photographed skewed, so the Short Duration Income units were read twice at 93% and wait for a person to confirm. A handwritten “p.a.” on the Zürich fee page reads at 91%. Kestrel Biosciences shows no cost basis at all, so the Statement Reader drafts a request to Lorraine, sent from Ethan through the client portal.
The Statement Reader never uses a value under 95% confidence and never moves on if a total does not reconcile.
Her 17 holdings sit on the left, X-rayed for cost per year, overlap and concentration; the proposed Quillfield Balanced Growth account sits on the right. The strip between them states the difference in plain terms: $14,880 a year less in total cost, all-in 1.55% → 0.92%, fund expenses 0.88% → 0.41%, risk score 68 → 54 inside her band of 48–58, overlap 71% → 18%, and her largest single stock, Kestrel, 9.8% → 4.0%.
“Cited: 3 statements, 13 fee tables, the risk model.”
Ethan opens Harborline Large Cap Growth Fund A: $198,400 at a 0.98% expense ratio, $1,944 a year. The fee table comes from the prospectus dated March 1, 2026, page 4, for the exact share class: a 0.62% management fee, a 0.25% 12b-1 fee and 0.11% of other expenses; its maximum sales charge is 5.75%, paid at purchase. It overlaps 72% with Ardent Focused Growth Fund C, which she holds in a C share class at 1.79%.
The Cost Lookup agent never uses a fee older than the latest filing and never infers a share class it cannot see.
Below the X-ray the agents' findings arrive as a conversation. The Transition & Tax agent has handed the two Swiss equity funds, likely PFICs for a US person, to Grace Whitfield on the Tax & Planning desk. The Risk Fit agent notes that Lorraine says she would sell after a 20% fall, yet held through 2022. Ethan asks what he should raise on Thursday, and gets three questions with their sources, saved to the meeting notes in the CRM.
Transition & Tax: “I have not planned their sale. The Tax & Planning desk will check with her CPA whether an election was made.”
One switch recalculates the strip for each alternative: keep as is ($36,776, 1.55%, risk 68); keep and convert the C shares ($35,182, 1.48%); Quillfield Core Index 60/40 ($14,508, 0.61%, risk 55); a brokerage account with the same funds ($10,874, 0.46%); and the recommended Balanced Growth ($21,896, 0.92%, risk 54). The cheaper ones are not hidden; each carries the reason it was not chosen.
Core Index 60/40: “Lower cost. Not chosen: no tax-exempt bond sleeve for her two taxable accounts and no active downside management, which she asked for (questionnaire Q6, Q8).”
Her questionnaire of October 14, 2026, scores 54, band 48–58, moderate. Today's portfolio scores 68, ten points above, mostly because of 77.9% in equities and 9.8% in one stock; the proposal scores 54. In a 2022-style stress test today's portfolio loses 21.4% ($507,500) against 14.1% ($334,400) proposed, labeled hypothetical. Without the conflicting answer her score would be 57, still inside the band.
The Risk Fit agent never changes a client's answer; outside the band always escalates.
The Marketing Review agent checked the draft against the firm's rule library. A selective claim that a fund “has beaten its benchmark in 8 of the last 10 years” became the standardized 1-, 5- and 10-year net returns; “guaranteed lower costs” became “lower expected costs at current fee schedules”. The line “Over ten years, the savings could grow to $187,200” stays blocked until Ethan adds disclosure HYP-04 with its assumptions. Then he approves v4 as maker, with a reason.
“You approve proposal PRO-26-10418 v4 for Lorraine Vance. It goes to Robert Delgado, registered principal … You cannot review it as principal.”
Robert's queue shows 24 proposals waiting in the New York branch on a one-business-day clock. For Lorraine's, the agents have prepared the evidence: Form CRS delivered on October 14 before the recommendation, statements reconciled, costs cited to 13 fee tables and 3 account fees, 5 alternatives with reasons, risk fit, the IRA transfer documentation under PTE 2020-02, the marketing lines cleared and the PFIC exception. He approves with a condition; only then can Ethan deliver it.
Robert's reason, on the audit trail: “Best interest supported; condition: PFIC tax review before Zürich sale.”
The same proposal, seen by four of the people who carry it — what their week looked like, and what it looks like now.
Seven agents read, price, X-ray, test and draft every proposal; the advisor approves and a registered principal reviews.
Reads brokerage, IRA and bank statements from any custodian — PDF, scan or photo — in the original language; extracts positions, units, prices, cost basis and fees, converts foreign currency at a cited closing rate and reconciles every statement to its total.
Finds each holding's fee table in its own prospectus or KID for the exact share class and pins it to the page, then adds the account-level wrap, custody and IRA fees from the statements to give the all-in cost.
Looks through fund holdings for overlap, sector and single-stock concentration — the large-cap funds that own the same stocks, and the names that appear in all of them.
Scores today's and the proposed portfolio against the risk questionnaire band, runs labeled hypothetical stress tests and checks the client's answers against her own statement history.
Estimates realized gains, sales charges and surrender charges, phases sales across tax years, and stops on PFIC and illiquid holdings.
Drafts the proposal in plain words with the cost-and-alternatives comparison, the account-type rationale and the transition plan, from the same cited numbers.
Checks performance, hypothetical and promissory language against the firm's rule library under the SEC Marketing Rule and FINRA Rule 2210, proposes fixes and pairs lines with approved disclosures.
Approve, review, clear. The agents propose; a named person decides.
Ask in plain words, or tell the agents what to do. They answer with citations, act for you, and everything they do is logged as theirs.
Why is Lorraine Vance's all-in cost 1.55%?
Four parts, all from her own documents: fund expenses $20,870 (13 fee tables, weighted 0.88%); the Harborline wrap fee $13,104; the Zürich custody fee $1,810, written in German; and IRA and trade fees $992. Total $36,776 on $2,371,640.
What should Ethan ask on Thursday?
Three questions: (1) In 2022 you held through a 20% fall; would you really sell next time? (2) Can you send the cost basis for Kestrel? (3) Did your CPA make an election for the Aldrich funds? I saved them to the meeting notes in the CRM.
Add the C-share conversion to Lorraine's alternatives
Done: alternative 2 converts Ardent Focused Growth C (1.79%) to A shares (1.04%), $1,594 a year less, no tax and no sales charge. It is in the alternatives table for the principal.
Which proposals are past the principal SLA?
One in the New York branch: Okonjo Family Trust (PRO-26-10366, Samuel Pryce), overdue by 25 minutes; it has a rollover that needs the PTE 2020-02 analysis. Hollis Bramwell is due in 1 h 10 m. The branch manager is copied at 1.5 days.
13 screens from the working solution, on its sample data. Pick one to see it large.
The thresholds every proposal is checked against are settings. Each is versioned, and a change takes effect only after a second approver — another administrator or the Chief Compliance Officer — agrees.
| Setting | Default | Choose from |
|---|---|---|
| Confidence needed to use an extracted value | 95% | Below it, the run stops and asks a person |
| Risk band tolerance | 0 points | Outside the questionnaire band always escalates to the principal |
| Cheaper alternative | Any | Every cheaper reasonable alternative needs a written reason |
| Single-stock limit after transition | 5% | Firm concentration policy |
| Interval fund and illiquid limit | 10% | Liquidity policy |
| Prospectus age | Latest filed | A newer filing triggers a refresh before use |
| Marketing review by a person | Rules marked “Reviewer” + 10% sample | Everything else the agent clears when paired with approved wording |
| Principal review clock | One business day | Reminder to the principal at 0.75 days; branch manager and backup principal at 1.5 days |
Every figure is labelled: a target the solution is built to, an estimate, a typical published result, or a proven one.
“demo” = seen in the working solution, on its sample data · “target” = the design goal, measured in the live solution · “estimated” = our estimate · “typical” = published figures (an advisor survey on prospect proposals, 2015, reported by WealthBriefing — 48% of advisors spend two hours or more on one, 84% more than an hour; Morningstar 2026 US Fund Fee Study — asset-weighted average expense ratio 0.32% in 2025). Regulations are named as context: Reg BI (17 CFR 240.15l-1), Form CRS, DOL PTE 2020-02, SEC Marketing Rule 206(4)-1, FINRA Rule 2210, SEA Rule 17a-4. People, firms, funds and accounts named on this page are fictional — characters and sample data in the working solution.
An investment proposal generator, also called investment proposal software or an investment proposal tool, turns a prospect's statements into a proposal an advisor can present: today's holdings and costs, a recommended portfolio and the comparison between them. Investment Proposal Generator does it with seven AI agents that read held-away statements from any custodian, price every holding from its own prospectus, X-ray overlap and concentration, test risk fit and draft a cited fee and risk proposal with every alternative on file. The advisor approves and a registered principal reviews before the client sees it.
Yes. The Statement Reader works as a statement scanner for prospect statements: it reads PDFs, scans and photos in the original language, converts foreign currency at a cited closing rate and reconciles each statement to its total. In the working solution a German statement from a Zürich private bank is shown beside the English. Any value under 95% confidence stops and waits for a person.
From each holding's own prospectus or KID, for the exact share class, pinned to its page; the cost library holds 18,400 products and refreshes a fee table when a newer prospectus is filed. Wrap, custody and IRA fees come from the statements, so the all-in cost includes what sits outside the funds. It also catches a C share that could convert to a cheaper class.
The Portfolio X-ray agent looks through each fund's holdings and shows how much of each pair of funds' equity is the same stocks, the names that appear in most of them and the largest single-stock positions. It says so when the fund holdings are older than a quarter.
It builds the record a principal needs to judge the recommendation: a cost-and-alternatives comparison from the same cited numbers, each cheaper alternative with a written reason, the account-type rationale, the IRA transfer documentation under PTE 2020-02 and the date Form CRS was delivered. The registered principal reviews that evidence and decides.
The Marketing Review agent checks every draft against your firm's versioned rule library for hypothetical, performance and promissory language. It never clears a hypothetical line without its approved disclosure and assumptions, proposes standardized net returns in place of selective claims, and sends what it cannot clear, plus a 10% sample, to your marketing compliance reviewer.
Yes. The agents draft, cite and escalate; they never send anything to a client, never give tax advice and never plan a sale they cannot assess. The advisor approves as maker with a reason, a registered principal approves, returns or rejects with a reason, and delivery is locked until both have signed.
The Agentic Solution Engine builds and deploys it from your requirements and documents — your fee schedule, model portfolios, risk questionnaire, rule library and a sample of past prospect statements — and it goes live once every quality gate has passed. We will walk you through it on your own prospects first.
Accounts open the same day, and not-in-good-order stops before the custodian sees it.
1-min demoSource of Wealth VerificationEvery dollar of a client's wealth traced to cited evidence before the BSA/AML Officer signs.
1-min demoReg BI Compliance ReviewEvery recommendation reviewed overnight; every flag decided by a principal, with evidence.
1-min demoWe'll run Investment Proposal Generator on a sample of your own prospect statements, fee schedule and models.