Define the question before taking a view
The Jun Yuan agent does not begin with a directional recommendation. It first defines the market question and the primary variable: growth, inflation, the central bank's reaction function, term premium, or the rate differential and policy constraints behind an exchange rate.
Macro trading rewards changes at the margin, not conviction for its own sake. Start with what the market already prices in, then identify the expectation gap. If the anchor is wrong, direction, sizing, and risk limits quickly lose their discipline.
-
01
Define the variable
First ask whether this is a rates question, an FX question, a credit question, or a change in the central bank's reaction function.
-
02
Find the anchor
For rates, watch OIS, the curve, and term premium; for FX, watch rate differentials, carry, capital flows, and policy limits.
-
03
Place it in historical structure
Calibrate the mechanism against big dislocations like 2008, the euro crisis, and the 811 reform — without mechanically copying their conclusions.
-
04
Spell out triggers and invalidation
Every view must state which signal confirms it and which signal invalidates it.
How a macro desk frames the question
Ask about the Treasury market and it will not stop at recession odds. It breaks the question down: how much of the policy path has OIS already priced, whether the long end is responding to real rates, term premium, or fiscal supply, and whether the curve is consistent with growth expectations. If the anchors conflict, the market has not settled on a clear view.
Ask it things like
Best for questions about rates, FX, central-bank expectations, and macro risk. Start with the market move; the agent identifies the relevant anchors and the variable currently driving pricing.
What the output includes
The output follows a desk-style structure: define the question, identify the moving anchor, compare it with history or the central bank's reaction function, and specify the conditions to watch. For the yen, the relevant inputs include the US-Japan rate differential, carry, finance-ministry and central-bank communication, the political case for intervention, and the market's capacity to absorb it.
If you need current OIS, yields, exchange rates, the dot plot, or option-implied volatility, it won't quote numbers from memory. Exact figures must come from the latest data or from you; without current numbers, it tells you which anchors to watch rather than pretending to know real-time prices.
Who it's for
- Investors tracking Treasuries, the dollar, the yen, OIS, IRS, and central-bank policy expectations
- Anyone who wants macro narratives reduced to pricing anchors, trigger conditions, and risk scenarios
- Anyone curious how historical structures like 2008, the euro crisis, and the 811 reform map onto today
- Anyone who wants to define invalidation and downside scenarios before taking portfolio exposure
Boundaries: research framework, not a trading instruction
It does not provide buy/sell, allocation, or leverage instructions, guarantee returns, predict exact timing or levels, or claim to represent an actual position. For high-risk strategies—including leveraged trades, uncovered options, or directional attacks on managed exchange rates—it starts with assumptions, constraints, and downside risks.
When the evidence provides no reliable anchor, the agent says so. In macro research, forcing a directional view can be more dangerous than remaining undecided.

Jun Yuan
Bring a rates, FX, or central-bank question. The agent will identify the pricing anchor, compare it with history, and define the triggers and risks.
Built from public materials to demonstrate a Jun Yuan-inspired macro research framework. It does not represent Jun Yuan or provide investment advice.



