Argosy research
What can
investors own?
Some investments depend on who can participate, on which terms and with what authority. Our research examines those conditions.
The Organization of the Investment Frontier
Portfolio choice asks what an investor should hold. Shared investments raise a prior question: which opportunities can form? This paper studies investments that require several independently governed investors to participate. Better alternatives can make acting separately more attractive while removing a shared investment that remains jointly valuable. The paper identifies when that happens, when acceptable allocations already support an equilibrium, and when transfers, assurance or dependable funding can preserve an opportunity.
Personal mandates. Shared investment.
Which differences can coexist in a shared investment, and what must happen between attractive demand and completed investment?
Your portfolio is the brief.
How existing holdings, desired exposure and obligations guide sourcing, including ordinary routes and decisions to wait.
Invest together. Keep your own portfolio.
Shared implementation, different participation amounts, clear authority and continuing commitments.
Separate Records,
Shared Processing
An Experiment in AI-Assisted Investment State Maintenance
A paired experiment processes the same synthetic investment histories separately and together, with and without persistent records and validation. Records reduce repeated retrieval. Shared processing has lower recorded inference costs, while the tested improvement saves more in separate processing. Successful primary contexts are at the scored interpretation ceiling, and most exact-output improvement is a change in how abstention is represented. The report preserves failures, costs, prompts and receipts. It measures a component of processing cost, not investment returns or the total value of an investment office.
Download the methods reportHow to read the research
Foundations introduces the questions. The working paper states its models, assumptions and results. The methods report examines a bounded processing task using synthetic histories.
None is an investment performance record. The continuing Argosy office is a product vision informed by this work.
Previous working-paper edition · Refinement-3