A plain-English guide to central bank FX reserves.

When a finance ministry announces it is "examining ways to improve management of its FX reserves war chest", as Japan did this morning in the draft growth-strategy report we covered in today's analysis, the headline number gets quoted as if it were the operational firepower available for intervention. It is not. Japan's $1.3 trillion in published FX reserves contains roughly $50 to $100 billion that could actually be deployed in a single trading day. The rest is locked in longer-duration assets, foreign holdings, and gold. This is the explainer on what FX reserves actually are, how they are composed, and what the operational war chest really constrains.

What FX reserves actually are

Foreign exchange reserves are external assets held by a central bank or sovereign authority that are denominated in a foreign currency and available for balance-of-payments purposes. The IMF Special Data Dissemination Standard defines five reserve-asset categories:

• Monetary gold. Physical bullion held by the authority, valued at the LBMA fix.
• Special Drawing Rights (SDRs). The IMF's international reserve asset, a claim on a basket of major currencies.
• Reserve position in the IMF. The country's drawing position at the Fund.
• Foreign currency holdings. The largest category for most authorities, including Treasury securities, agency MBS, supranational bonds, deposits with foreign central banks, and bank deposits.
• Other reserve assets. Repurchase agreements and other instruments not captured by the first four categories.

Japan's $1.3 trillion in context

Japan's FX reserves are the second-largest sovereign FX reserve holdings in the world, behind China's roughly $3.2 trillion. The Japanese reserves are managed by the Ministry of Finance through the Foreign Exchange Fund Special Account (FEFSA), which is operationally distinct from the Bank of Japan's balance sheet. Three structural facts about Japan's holdings:

1. USD-denominated assets dominate. Roughly 75 percent of Japan's reserves are held in US dollar instruments. The breakdown is approximately 60 percent in long-duration US Treasuries, 15 percent in shorter-duration Treasuries, 10 percent in agency MBS and supranational bonds, 10 percent in gold and SDRs, and 5 percent in actually-liquid USD deposits.
2. Custody is at the Federal Reserve. The US Treasury holdings sit in a custody account at the Federal Reserve Bank of New York. Operational liquidity requires either selling Treasuries (which moves the Treasury market) or drawing on standing repo facilities.
3. The "war chest" framing overstates deployability. Selling $300 billion of long Treasuries to fund a yen-buying operation would itself move the US 10-year yield by an estimated 20 to 40 basis points, undermining the operation through the rate-differential channel. The operational maximum, without disruptive Treasury sales, is closer to $50 to $100 billion within a single trading day.

The composition picture

What previous interventions actually cost

The April 30, 2026 operation (the most recent before this week's verbal escalation) cost roughly $62 billion in reserves, which produced a peak USD/JPY drop of about 4 percent that fully reversed within six weeks. The 2024 April-May operation cost $73 billion for a 5.6 percent drop in reserves and a similar reversal timeline. The 2022 September operation cost $42 billion for a 3 percent move.

Three observations from the operation history:

• The cost-per-figure of USD/JPY decline is structurally rising. The 2022 episode cost roughly $14 billion per figure of USD/JPY decline; the 2024 episode cost $14.6 billion; the 2026 April episode cost $15.5 billion. The trend reflects deeper market depth and structurally wider rate differentials.
• The reversal timeline is consistent. Six to eight weeks. The signalling channel works but decays; without coordinated central-bank rate action, the pair returns to where the carry-trade math implies.
• The maximum operational size is constrained. A single operation has never exceeded $73 billion in the 2022-2026 sample. The published $1.3 trillion in reserves is misleading as a firepower indicator; the practical maximum per episode is closer to one tenth of that.

Why the war-chest framing matters anyway

A larger reserve stock does provide three real benefits even when the operational firepower is far smaller than the headline number:

1. Credibility through optionality. Markets cannot know in advance how much of the reserve stock the authority is willing to deploy. The expectation that the authority could theoretically commit a multiple of any previous operation creates option-side risk that compresses yen carry trades' implied volatility.
2. Coordination with monetary policy. Reserve management is operationally separate from Bank of Japan policy, but the two are coordinated through MoF-BoJ communication. The reserve stock can be used to support BoJ rate decisions through coordinated FX action.
3. Credit-rating support. Sovereign credit ratings incorporate FX reserve coverage. A larger reserve stock supports the credit rating and indirectly the cost of all Japanese sovereign and quasi-sovereign borrowing.

What the war chest does not tell you

1. The political authority to deploy it. The reserves are owned by the FEFSA, controlled by MoF. Deployment requires political approval and coordination with the Cabinet. A finance minister cannot unilaterally deploy reserves at the scale of the headline number.
2. The strategic intent. Reserve growth can reflect intervention-buying (selling yen), official-bond accumulation, or current-account surpluses. The same headline number can be reached through different strategic stances.
3. The G7 coordination availability. Japanese reserve deployment is most effective when paired with G7-coordinated action. Without US Treasury sign-off, the practical operation is solo and the effectiveness is bounded.

Related reading

• Today's analysis applies this directly: Dollar extends. Japan looks at its war chest.
• The intervention companion: A plain-English guide to FX intervention
• The carry trade context: A plain-English guide to the carry trade
• The real-yield differential driving the carry: A plain-English guide to real yields