Skip to content
Directory
GitHub Star
Evidence, tags & source code

More Than One Investment Expert: How Multi-Perspective Agents Test an Investment Thesis ​

Multi-Agent Investment Research · Part Three | Shared facts, different methods, and disagreements worth resolving

The value of multiple investment experts is not a vote. It is an opportunity to falsify a claim that could materially affect an investor's decision. Agreement is not independent confirmation if every specialist relies on the same incomplete material.

Suppose an investor already holds several funds and wants to add a technology ETF because “more funds means better diversification.” The useful questions are whether underlying exposures overlap, whether liquidity constraints permit the change, and what evidence could overturn the diversification claim.

We call the model responsible for the user’s question and final synthesis the lead researcher (Main), and an independently executed, delegated task a research assistant (Child). Bull and Bear support and challenge a particular hypothesis; they do not vote on the result.

That is why ETF Agent uses optional investment methods such as Markowitz and Taleb alongside Bull/Bear challenge. A Child is worth the additional calls and delay only when a different method can test something material. Main remains responsible for a qualified, evidence-based synthesis. Peer-message and recovery mechanics belong to the Bull/Bear Runtime article; here the focus is choosing and using different perspectives.

1. Investors Do Not Simply Need More Experts ​

“What is this ETF?” can often be answered using its benchmark, fees, holdings and risk disclosures. “Should I add it to my portfolio?” also requires information about existing exposures, the funding source and the investor's longer-term plan.

No single metric answers all those questions. High growth does not imply a low valuation; owning more fund names does not ensure less underlying concentration; lower historical volatility does not exclude severe losses.

We want multi-perspective research to do three things: identify important assumptions, apply suitable methods to the same underlying facts, and revise a reason when legitimate counterevidence emerges—rather than vote on competing recommendations.

That means specialists are optional. If the only question is a fund's current management fee, a permitted disclosure lookup may be sufficient. An extra Child Agent is justified only when different analytical methods or an independent challenge might add material value relative to its cost in time and model calls.

2. What Do the Six Investment Methods Actually Examine? ​

We registered six optional investment methods on the existing Child execution mechanism. Their names refer to analytical schools and established research traditions. The model does not impersonate these people or invent their current investment advice.

PerspectiveIts first questionWhat it cannot conclude from the label alone
Markowitz: allocationHow do correlations, marginal risk and risk contributions change?Without aligned return series, sector names cannot yield an optimal allocation
Taleb: tails and fragilityWhat happens under market, liquidity or cross-border infrastructure stress?Historical VaR does not capture every black swan or justify fabricated probabilities
Bernstein: long-term disciplineIs the change consistent with the time horizon, cash needs and plan?Current holdings do not establish the investor's true risk tolerance
Bogle: implementation costWhich index, tracking difference, overlap and total costs matter?The lowest displayed price is not always the most suitable implementation
Marks: compensation for riskDoes today's price adequately reward the risks being taken?A macro story cannot precisely forecast a market turning point
Fama–French: factor explanationWhich comparable historical factors may explain observed returns?An explanatory regression is not manager skill, causality or a forecast of alpha

For a proposed technology ETF, Markowitz may focus on correlation with current holdings; Bernstein may question whether cash earmarked for a payment six months from now is truly available; Bogle may compare the total cost of alternative funds offering the same intended exposure. They need not arrive at identical sentences because they are answering different questions.

The value for the investor is not that “six professionals appeared.” It is that one decision is tested against different, clearly identified criteria.

3. Independence Starts With Shared Facts, Not Shared Opinions ​

Generating three separate reports is easy. The hard part is avoiding different disclosure dates, incompatible data definitions and seemingly irreconcilable conclusions.

Our design prefers to reuse the current Run's authorized source material and deterministic calculations that Main already has. Existing inputRefs and Result/Lineage identities can pass actual result references to a Child. A new expert should not automatically fetch the same portfolio again or maintain its own competing truth database.

Three concepts must remain separate:

  1. Fact: dated holdings, a product disclosure or an actually executed calculation, with its source and measurement basis.
  2. Hypothesis: a testable proposition such as “adding this ETF improves diversification.”
  3. Opinion: the interpretation a particular investment method develops from those facts and hypotheses.

If Main tells every Child that it already believes the purchase improves diversification, and smuggles that provisional belief into the source facts, “independent analysis” may simply repeat Main's initial position. A better handoff carries the original materials and user restrictions. When critiquing an earlier view, label that view as a claim to be examined, not verified evidence.

Sharing result identities also does not share permissions. A Child authorized to read protected data cannot grant sibling Children access by paraphrasing it. Host authorization and resource rules still govern every real tool call, disclosure and action.

4. What Should Several Agents Do With the Same Portfolio? ​

We use entirely synthetic teaching data, not real holdings. Suppose the original portfolio contains a broad equity fund (50%), bonds (30%), and cash (20%). The broad fund is 30% technology by the chosen classification, while the candidate technology ETF is 80% technology. We assume the disclosure dates, classifications and coverage are aligned and complete; bonds and cash have no equity technology exposure under this teaching definition.

PlanBroad equity fundTechnology ETFBondsCashTechnology-sector exposure
A: do nothing50%0%30%20%15%
B: buy with cash50%10%30%10%23%
C: replace broad equity40%10%30%20%20%

The deterministic calculation is simply the sum of each fund's portfolio weight multiplied by its technology-sector weight. Three experts need not each recalculate it. The interesting question is how they use the same qualified number differently.

  • Allocation perspective: Plan B lifts technology-sector exposure from 15% to 23%, raises the overall equity allocation and reduces cash. Without comparable return and covariance data, it still cannot prove that overall portfolio volatility must increase.
  • Long-term discipline: Does that reduction in cash use money needed for a near-term payment? Until the investor confirms the need, the answer remains unknown.
  • Implementation perspective: If the real objective is an intentional technology overweight, which ETF benchmark, fees, trading conditions and overlapping holdings justify the implementation? Missing actual fee data cannot be replaced with precise-sounding numbers.

These observations challenge a specific proposition: “Because there is one more fund, technology exposure cannot rise.” They do not establish that all measures of diversification worsened, or that the investor must buy or sell.

This is a common multi-Agent failure mode: concatenate three expert reports and call the result “synthesis.” Genuine synthesis relates different methods to one choice, including conditions that are still unresolved.

5. When Is a Bull–Bear Exchange Worthwhile? ​

Different methods do not always require a debate. For “What is the overlap percentage?”, source retrieval and calculation matter most. When an important decision depends on a falsifiable hypothesis, an independent challenger may become valuable.

Suppose original hypothesis H1 says: “Plan B does not increase technology-sector exposure.” Its original falsification condition is: on complete, comparable look-through data, B's technology weight is greater than A's.

A simplified synthetic teaching exchange, not a transcript from a real model run, might look like this:

Bull: The technology ETF might increase targeted growth exposure, but I have not verified the assertion that more funds necessarily improve diversification.

Bear: Under the aligned data, Plan B increases technology exposure from 15% to 23%, satisfying H1's original falsification condition. Should you withdraw that specific proposition?

Bull: I withdraw H1. If the investor wants an intentional technology overweight, we still need to assess the price, cash needs and whole-portfolio risk. We cannot claim that improved diversification was already demonstrated.

The useful change is not that Bear “won.” Bull actually revised a clearly identified reason.

ETF Agent retains ordinary Main-to-Child delegation and Child-to-Main result handoff. An admitted Bull/Bear pair may directly exchange research challenges and replies within the same Run, without Main relaying every message. The Host checks that shared prose and source references are permitted for the recipient; hidden reasoning and inaccessible source material cannot be passed through. The messages remain available through authorized History views, not as public Internet posts. Either peer may stop early rather than exhaust a round quota. The Bull/Bear runtime article covers message delivery, waiting and recovery.

6. Main Is Responsible for Synthesis, Not a Vote ​

If two experts disagree, which one should the investor trust?

A vote, averaged position or consensus score cannot replace a supported investment assessment. If one perspective needs correlations and another needs confirmation of cash requirements, averaging “buy 10%” and “do not buy” into a 5% order resolves neither missing input.

Main should bring together three categories: sourced facts and actual calculations; attributed findings and material differences from the specialists who truly participated; and conditions still unknown or awaiting investor clarification. If a Child fails or has not produced a result, Main cannot manufacture that specialist's vote. Independent valid findings remain deliverable.

A suitable user-facing synthesis, still entirely hypothetical, might say:

Verified change: under these complete synthetic inputs, Plan B takes technology exposure from 15% to 23% and cash from 20% to 10%.

Revised rationale: owning one additional fund did not establish that technology exposure was diversified. That specific reason should be withdrawn.

Unresolved: total portfolio volatility, actual cash needs, ETF fees and valuation inputs remain insufficiently established.

Next step: if the true goal is an intentional technology overweight, analyze that goal and its implementation directly. Do not create an automatic trading instruction.

The investor benefit is not that AI sounds more confident. It is knowing which reason changed and what remains necessary for a decision. Demonstrating a real quality improvement requires independent comparison, not this synthetic case alone.

Optional Research Reviewer: Audit the Study, Not the Vote ​

Investment perspectives determine which method to apply; Bull/Bear determine who challenges a specific premise. ETF Agent also implements an independent Research Reviewer Child to inspect the original evidence, material gaps, and whether further investigation is worth the cost. It is neither a seventh investing method nor a third peer in the Bull/Bear dialogue.

If Main is preparing to claim that another technology ETF improves diversification, Reviewer can check the original authorized portfolio facts, calculations, and disclosure dates—not just Main's summary. Has the analysis confused industry weight with total portfolio volatility? Did it overlook cash needed in six months? Reviewer may conduct authorized targeted verification and suggest continue, pivot, finalize, or suspend.

Reviewer reuses existing Child execution, budget, and recovery. Main can call consult_research_reviewer when manual review is enabled; Runtime may admit a bounded automatic review only under configured factual triggers and resource limits. Reviewer cannot approve an answer, certify investment truth, or block a lawful Main publication. Sufficiently supported research can be delivered without a review.

The path has scoped engineering acceptance for review execution and advice handoff; real financial-quality gains and incremental cost still need separate evaluation. There is also an explicit post-publication financial-semantic review report; this separate evaluation tool is not the in-Run Reviewer Child or a mandatory final approval step.

7. Under the Hood: Several Methods, One Child Runtime ​

We did not create six runtimes for six investment methods. New expert Children now select a registered method using the unified agentId, reusing existing Child tools, grants, budgets, input handoff, and recovery. Already sealed historical descriptors retain their original meaning; new parameters must not silently reinterpret them.

Independent analysis depends less on extra role fields than on three conditions:

  • Every expert method has a traceable version that cannot quietly change during historical recovery.
  • Main should pass the actual user question and originally authorized results, not promote its own provisional conclusion into source truth. Incomplete evidence may justify more research.
  • An expert returns findings, disagreements, and gaps. It does not acquire additional private sources, trading authority, or final publication power.

Read the relationships in groups; the complete diagram remains available below.

Investor asks a question
  • → Main confirms goal and available evidence
Main confirms goal and available evidence
  • → One research path sufficient? (decision)
One research path sufficient? (decision)
  • Yes → Main researches and delivers
  • No → Select relevant method and Child on demand
Select relevant method and Child on demand
  • → Children read the same authorized facts
Children read the same authorized facts
  • → Material hypothesis worth challenging? (decision)
Material hypothesis worth challenging? (decision)
  • No → Hand back findings and gaps
  • Yes → Optional Bull Bear falsification dialogue
Hand back findings and gaps
  • → Main synthesizes sources and unresolved conditions
Optional Bull Bear falsification dialogue
  • → Hand back findings and gaps
Main synthesizes sources and unresolved conditions
  • → Main researches and delivers
Complete relationship diagram · enlarge for detail
Agent architecture diagram
Agent architecture diagram 放大查看完整图 / Open full-size SVG

The diagram illustrates optional collaboration, not a mandatory workflow for every investment question. A simple fee lookup may not need an expert; a material hypothesis with genuinely different available methods might.

8. When Is an Extra Specialist Worth the Cost? ​

The six registered methods have scoped engineering evidence for registration, handoff, and recovery; Bull/Bear communication has separate execution evidence. This does not yet show that more expert Agents produce better investment decisions. Real models may still choose the wrong specialist, delegate repeatedly, or produce more prose without a useful counterexample.

Before adding a specialist, we now prefer three questions: Can ordinary Main already answer the task? Will a new perspective challenge an assumption that affects the decision? Can the new finding be checked against the same authorized evidence?

A management-fee lookup may need only one direct read. A claim that a technology ETF lowers concentration may benefit from different tests of correlation, tail risk, and liquidity discipline. Disagreement should become a conditional statement about what would change the decision, not a vote count.

A full benefit claim still needs comparable single-Main and multi-expert research on the same questions and data, including new falsifications, latency, and cost. Such financial-quality comparisons remain open.

The value of investment research is not how many experts agree. It is whether an important reason survives meaningful tests from different methods.


Further Reading ​

Subagent foundations explain context isolation and result handoff; bounded Bull/Bear dialogue covers direct messages, waiting and recovery; the 2026-10-24 historical collaboration outlook preserves the design hypothesis and comparison questions from that time, not current implementation guidance. For the history of research reasons, see Investment Thesis and continuing weekly research.

The examples are synthetic, not a real personal portfolio or investment advice. This essay explains the internal research system’s design; the public teaching implementation does not contain all of these capabilities. Its actual scope is described in the public repository README. More essays are in the article collection; Stars and constructive counterexamples are welcome.

Engineering notes, evidence and boundaries. · Contact: hpoenixf@foxmail.com · WeChat: hpoenixf0510