Financial Advisor Research: Workflows, Data & Signals

The most popular advice about financial advisor research is also the least useful: read more reports, add more data feeds, and keep up with every market narrative. That approach creates crowded inboxes, not better decisions. A defensible research process does something harder. It identifies which evidence deserves attention, tests it against contradictory information, and records why the final recommendation remains reasonable.

The need is growing. UK firms giving financial advice reported around £1 trillion in assets under advice and 4.1 million retail clients in 2025, while adviser headcount remained broadly stable at about 31,000. The number of authorised advice firms fell 15% since 2021, and large firms accounted for around 50% of assets under advice, according to the UK financial advice industry snapshot. In the United States, the Investment Adviser Association reported 16,544 advisers in 2025 and 73.7 million clients, with client numbers increasing 7.7%. McKinsey's advisory market analysis also describes fee-based advisory relationships growing from about $150 billion in 2015 to $260 billion in 2024, a 6.4% compound annual growth rate.

Those figures describe a larger research burden, not a license to outsource judgment. The practical question is whether a recommendation rests on independent, current, and incentive-aware evidence, or whether it's a home-office model portfolio wrapped in a client conversation.

Why Most Advisor Research Is Broken

Most advisors don't perform independent security research in the way clients imagine. A 2025 survey found that 44% of advisors used home-office model portfolios and recommended lists, 37% used wholesaler recommendations, about 15% used subscription research platforms, and 13% used internal research, as reported in Nasdaq's survey-based discussion of advisor differentiation.

That doesn't make centralized research useless. Home-office teams can provide governance, manager due diligence, trading support, and a consistent investment policy. The problem starts when a platform recommendation becomes the entire basis for conviction. A wholesaler's deck may summarize a fund's positioning, performance history, and portfolio commentary, but it's still a sales document with a carefully chosen frame.

Centralization creates consistency, not necessarily insight

A model portfolio can make implementation easier, yet it can also produce portfolios that look alike across firms. If several practices use the same approved list, the advisor's contribution may shift from investment selection to planning, behavioral coaching, tax coordination, and communication. That's a legitimate service model, but it shouldn't be presented as bottom-up research if the advisor hasn't independently tested the underlying holdings.

Conflicts deserve equal attention. Revenue-sharing arrangements, distribution incentives, and product economics can influence what reaches an advisor's desk. A recommendation may still be suitable, but suitability alone doesn't answer whether the process served the client's best interest or whether cheaper, simpler alternatives received a fair comparison.

The incentive problem isn't theoretical. In a controlled NBER audit study of financial advice, advisers didn't systematically correct client mistakes and often reinforced profitable behaviors, including return-chasing and recommendations toward higher-fee active funds when clients began with diversified, low-fee portfolios. That finding makes advisor output a noisy signal, not a final verdict.

Practical rule: Treat firm-issued research as an input. Require an independent check of fees, portfolio exposure, incentives, and thesis risk before presenting a recommendation as your own.

Regulation also changes the observable market. An NBER study on fiduciary duty and market structure found that imposing fiduciary duty reduced the total number of firms in a market by about 9% and changed the composition of the firms that remained. Advice channel and regulatory regime therefore affect the recommendation set itself. Comparing advisor behavior across markets without considering those differences can lead to false conclusions.

The Anatomy of a Research Workflow

A sound workflow is repeatable without becoming mechanical. It should tell an advisor how an idea enters the system, what evidence can eliminate it, how conviction is recorded, and which developments force a review.

Start with a broad but controlled idea funnel

Idea generation can come from a market screen, an economic theme, a client constraint, a sector review, or an unusual insider activity alert. The source matters less than the next question: what would make this idea worth researching rather than merely interesting?

A preliminary screen should narrow the universe before anyone spends hours reading filings. Useful filters include valuation, earnings revisions, balance-sheet resilience, cash generation, liquidity, and portfolio fit. Technical signals can help with timing, but they shouldn't substitute for understanding the business.

Move from numbers to business reality

The deep dive begins with primary documents. Read the latest 10-K, 10-Q, 8-K, and proxy statement where relevant. Track revenue drivers, customer concentration, margin structure, capital allocation, debt maturities, share dilution, and management incentives. Earnings calls add tone and context, but management commentary needs to be compared with reported results and prior guidance.

Sell-side estimates can provide a useful map of expectations. They shouldn't become your model by default. Build an independent view of revenue, margins, cash flow, and balance-sheet requirements, then identify which assumptions explain the difference from consensus.

Force a decision before capital is committed

A conviction score doesn't need false precision. It needs consistent questions:

1. Business quality: What does the company do better than competitors, and how durable is that advantage?
2. Financial evidence: Are growth, margins, cash flow, and debt moving in a direction that supports the thesis?
3. Valuation: What expectations are already reflected in the price?
4. Catalysts: What could change investor perception?
5. Risks: Which facts would invalidate the thesis?
6. Portfolio role: What exposure does the position add, and how does it interact with existing holdings?

Write the thesis in plain language before buying. Include a base case, a bear case, and explicit review triggers. A position should be reduced or removed because the evidence changed, not because a new headline was emotionally persuasive.

Ongoing monitoring is where many practices weaken. Assign ownership, set review dates, and track the original assumptions alongside new information. If the company misses a target but the underlying economics improve, that's different from a business that beats estimates by cutting investment and weakening its future cash generation.

Traditional Data Sources Versus Alternative Signals

Traditional sources remain essential because they provide different kinds of evidence. None should carry the entire thesis.

Data Source	Type	Timeliness	Objectivity	Signal Strength
Sell-side research	Analyst interpretation and estimates	Often current, but dependent on publication cycles	Moderate, with potential commercial conflicts	Useful for expectations and debate, weaker as independent proof
SEC filings	Reported financial and legal information	Authoritative, but generally backward-looking	High for disclosed facts, limited for interpretation	Strong foundation for fundamental diligence
Earnings call transcripts	Management commentary and analyst questioning	Timely around results	Mixed, because management controls much of the narrative	Valuable when compared with prior statements and filings
Form 4 insider filings	Reported insider transactions	Often close to the transaction date	High for the transaction itself, limited for motive	Stronger when purchases cluster and context supports them
13F filings	Institutional ownership disclosures	Delayed relative to the underlying activity	High for reported holdings, limited for current positioning	Useful for ownership direction, not short-term timing
Short interest	Market positioning data	Periodic	Objective as a measure of disclosed short exposure	Helpful for understanding pressure and disagreement
Options flow	Derivatives activity	Potentially timely	Objective transaction data, ambiguous intent	Contextual, because hedges can resemble directional trades
Web or satellite metrics	Alternative operating indicators	Can be timely	Variable, depending on collection and interpretation	Potentially useful, but vulnerable to noise and sampling bias

Sell-side research earns a place in the workflow when it helps answer, “What does the market expect?” It's less reliable as the answer to, “What do I independently believe?” The report's target price, rating language, and scenario analysis may reveal consensus assumptions, but they can also reflect institutional relationships and incentives that the client doesn't see.

Filings are slower, yet they're harder to replace. A 10-Q may not tell you what will happen next quarter, but it can expose working-capital stress, debt covenant pressure, stock-based compensation, or changes in customer economics that a polished presentation minimizes.

Why insider activity can add a useful validation layer

Insider transactions are most informative when the transaction is discretionary and economically meaningful to the insider. Open-market purchases by senior executives, particularly when several insiders buy within a short window, can provide a separate behavioral signal. They don't prove that a stock is undervalued, and they don't reveal the complete investment case. They do help answer whether executives are willing to commit personal capital under the conditions they describe publicly.

The distinction between a Form 4 purchase and a scheduled sale matters. A routine 10b5-1 sale may reflect a prearranged diversification plan rather than a negative view of the business. An open-market purchase requires a different interpretation, but still needs to be checked against liquidity, compensation, ownership concentration, and corporate events.

Use alternative data to challenge the thesis, not decorate it. A signal that agrees with every other input may be reassuring. A signal that conflicts with the fundamentals is often more valuable because it tells you where the research requires another level of scrutiny.

Analytical Techniques That Separate Signal From Noise

Advisors don't need every available data point. They need a method for deciding which data points change the probability of the thesis.

Use signal stacking, not single-indicator trading

A single insider purchase usually has limited meaning. The transaction may be small relative to the executive's wealth, connected to compensation, or timed around a corporate event. The signal becomes more interesting when it aligns with independent evidence, such as improving cash conversion, stabilizing debt levels, credible earnings revisions, or a valuation that doesn't require heroic assumptions.

A practical stack might combine:

• Insider activity: Look for discretionary open-market purchases, multiple participating executives, repeated accumulation, or a first purchase after a long period without buying.
• Fundamentals: Compare the signal with free-cash-flow yield, gross and operating margins, debt-to-equity movement, working capital, and return on invested capital.
• Ownership: Review institutional changes through 13F filings, while remembering that the disclosure is delayed and may not reflect current positioning.
• Derivatives: Investigate unusual call or put activity, but test whether the trades could represent hedges, spreads, or broader portfolio positioning.
• Price and expectations: Ask whether the market already reflects the improvement suggested by the signal.

The point isn't to count agreeing indicators. It's to map whether each indicator has a different causal relationship to the thesis.

Read the transaction before interpreting it

Start with the filing details. Identify the insider's role, transaction code, price, number of shares, resulting ownership, and whether the trade was open-market or part of a plan. Then compare the purchase with the insider's prior activity and compensation structure.

Next, challenge your own interpretation. Would you still like the company if the insider signal disappeared? What evidence would make you reject the position? Does the thesis depend on a single management promise, one valuation multiple, or an earnings event that could disappoint?

A useful signal should narrow uncertainty, not merely increase excitement.

A simple scoring rubric can prevent confirmation bias. Score business quality, financial trend, valuation, insider context, market expectations, and downside risk separately. Record both supporting and opposing evidence. If insider buying looks constructive but debt is rising, customer concentration is worsening, or cash flow is deteriorating, the contradiction belongs in the recommendation memo.

Recency bias creates another trap. A fresh alert feels more important than a filing trend that developed over several quarters. Give the thesis a defined time horizon and specify which events matter within that horizon. Monitoring then becomes a test of assumptions rather than a stream of reactions.

Compliance and Fiduciary Considerations in Research

Fiduciary duty governs the research process, not just the final recommendation. An advisor needs a reasonable basis for believing an investment is appropriate, but that basis should also be documented, reproducible, and consistent with the firm's stated methodology.

A home-office list can be part of that process. A wholesaler deck can provide background. Neither should be the only evidence supporting a client-specific recommendation, particularly when the product carries higher fees, complex risks, or distribution incentives.

Build an evidence trail

A compliance-friendly research file should show:

• The original question: What client need, portfolio gap, or market observation started the review?
• The sources used: Record filings, research reports, earnings transcripts, data services, and alert records.
• The conflicts considered: Note compensation arrangements, fund expenses, affiliations, ownership, and limitations in third-party research.
• The alternatives reviewed: Explain why comparable securities, funds, or strategies were rejected.
• The decision logic: Connect the evidence to the portfolio role, risk tolerance, time horizon, and client constraints.
• The review triggers: State what would cause a reassessment, trim, or exit.

Insider data requires a clear boundary between public information and material nonpublic information. Publicly reported Form 4 transactions can be reviewed as disclosed filings, but advisors must avoid soliciting, accepting, or acting on confidential information obtained through restricted channels. Compliance teams should define approved data providers, access controls, escalation procedures, and personal trading rules.

Make methodology understandable to clients

Form ADV disclosures and client-facing materials should describe the firm's research approach accurately. If the practice uses centralized models with independent checks, say that. If it uses alternative data as corroboration, explain the role and limits of that information without implying predictive certainty.

A defensible process can strengthen fiduciary practice because it shows how the advisor reached a conclusion and how the firm manages uncertainty. It also gives the chief compliance officer something concrete to test during reviews.

Real World Example of Insider Signals in Action

Consider an RIA reviewing a mid-cap industrial company already held in several client portfolios. An alert identifies the CFO and two division heads purchasing shares on the open market within a two-week window. That event deserves investigation, not an automatic allocation increase.

The advisor begins by saving the filing details and recording the alert in the research log. The team then pulls the latest 10-Q, checks debt maturities and working capital, compares free cash flow with reported earnings, and reviews the most recent earnings call for changes in management's tone around demand, pricing, and backlog.

The initial signal strengthens if the executives made discretionary purchases at prices near the market, the company's balance sheet remains sound, and management's operating commentary is consistent with the reported numbers. It weakens if the transactions are tied to compensation, if insiders bought immaterial amounts, or if the filings reveal deteriorating liquidity that management didn't address clearly.

Step	Action Taken	Data Source	Signal Outcome
Initial alert	Verify buyers, transaction codes, prices, and resulting ownership	Form 4 filings	Potentially constructive, pending context
Financial review	Test liquidity, leverage, margins, and cash conversion	Latest 10-Q and 10-K	Either supports or challenges the insider signal
Management review	Compare current commentary with prior guidance and reported results	Earnings call transcript	Measures consistency of the narrative
Ownership check	Review changes in institutional positions	13F filings	Adds context, with timing limitations
Derivatives review	Investigate unusual options activity and possible hedging	Options market data	Contextual evidence, not a standalone conclusion
Portfolio decision	Compare expected return, downside, sizing, and client suitability	Internal model and investment policy	Hold, add, trim, or reject
Recordkeeping	Preserve the thesis, evidence, conflicts, and review triggers	Research management system	Creates an auditable decision trail

Suppose institutional ownership is stable, options activity shows no clear directional edge, and the 10-Q confirms that cash generation remains healthy. The advisor might decide to add modestly to an existing position, subject to client mandates and portfolio concentration limits. If a subsequent earnings report exceeds the firm's expectations, that outcome doesn't prove the insider signal caused the result. The original decision remains defensible because the advisor used a documented mosaic of evidence rather than a prediction based on one alert.

The opposite example matters more. Several insiders appear to be selling, but the filings show scheduled 10b5-1 transactions and ordinary diversification. The advisor shouldn't treat those sales as a bearish thesis without additional evidence. A disciplined workflow prevents both errors, chasing a purchase and panicking over a planned sale.

Building a Better Research Process Starting Today

Improving advisor research doesn't require a large terminal budget or an enormous analyst team. It requires a smaller number of ideas, clearer ownership, and a record of how each decision was made.

Start with an audit of the current process. List every source used in security selection, then label each as primary evidence, external interpretation, sales material, or alternative signal. Identify where the firm relies on approved lists, model portfolios, consensus estimates, or wholesaler presentations without an independent challenge.

Add one independent validation layer

Choose a signal that answers a question your current process doesn't answer. For many firms, that could be public insider transactions. For others, it might be filing-based anomaly detection, institutional ownership changes, customer activity, or a more structured review of earnings-call language.

A workable cadence can be simple:

• Weekly holding scan: Review new Form 4 activity, material filings, estimate changes, and major thesis triggers affecting current positions.
• Monthly deep dive: Select one high-conviction holding or watchlist name and rebuild the thesis from primary documents.
• Quarterly thesis review: Reconcile actual results with the original model, record which assumptions changed, and decide whether the position still earns its place.
• Event-driven review: Reassess immediately after a management change, capital raise, debt event, major acquisition, or unusual insider activity.

The technology should support the workflow rather than replace judgment. A spreadsheet, research management system, or structured CRM note can capture the thesis, evidence, portfolio role, sizing rationale, and exit conditions. Alerts should reduce search time, but every meaningful alert still needs human review.

Measure process quality, not prediction theater

Track whether recommendations have complete source records, explicit downside cases, conflict checks, and scheduled review dates. Review rejected ideas as well as purchased ones. A good process should help the team explain why it said no, not merely defend what it owns.

Altymo is one option for this validation layer. Its AI-powered insider trading tracker scans SEC Form 4 filings and surfaces context around events such as CEO or CFO open-market purchases, cluster buying, repeated accumulation, and unusually large trades, with the underlying filings and transaction history available for review. Advisors can use that information to identify questions for deeper diligence, not to bypass suitability analysis or fiduciary judgment.

The standard should be better-filtered research, not more research. When every position has a documented thesis, an independent corroboration step, and clear invalidation triggers, client communication improves alongside investment governance. Treat that discipline as an operating habit, with regular reviews of the sources, assumptions, and compliance controls that support it.

---

Use Altymo to add structured SEC Form 4 monitoring to your financial advisor research workflow and investigate insider activity with its underlying filing context. Visit Altymo to see how its alerts can help your team find relevant signals faster, then validate every signal through your own fundamental, portfolio, and compliance process.