Chapter 7 7.6

Arbitrage Strategies

Simple arbitrage concepts in JGB markets

Arbitrage Strategies

In financial markets, arbitrage refers to the simultaneous purchase and sale of related securities to profit from price discrepancies, ideally with zero risk and zero net investment. In practice, true “risk-free” arbitrage is rare in the highly efficient JGB market. However, relative value arbitrage opportunities do exist, where small mispricings can be exploited with limited risk.

This section covers the main types of arbitrage and relative value strategies in JGB markets.

Types of Arbitrage in JGB Markets

1. Cash-Futures Arbitrage

This is closely related to the basis trades covered in Section 4.5. When the futures price deviates from its fair value relative to the cash CTD bond, arbitrageurs can:

When futures are overpriced:

  • Buy cash CTD bond
  • Sell futures
  • Hold to delivery for risk-free profit

When futures are underpriced:

  • Sell cash CTD bond (or short it)
  • Buy futures
  • Unwind or deliver at convergence

In reality, transaction costs and market efficiency mean this is rarely true arbitrage, but rather a low-risk relative value trade.

2. On-the-Run vs. Off-the-Run Arbitrage

This is one of the most common relative value opportunities in JGB markets.

On-the-run bonds are the most recently issued JGBs at each maturity (2Y, 5Y, 10Y, etc.). They typically trade at:

  • Lower yields (higher prices) than off-the-run bonds
  • Tighter bid-offer spreads
  • Higher liquidity

The on-the-run premium (or liquidity premium) is the yield difference:

\[\text{On-the-Run Premium} = Y_{off-the-run} - Y_{on-the-run}\]

Typical ranges:

  • 10Y JGBs: 2-5 basis points
  • 5Y JGBs: 1-3 basis points
  • 2Y JGBs: 0.5-2 basis points

The Trade:

  • Buy off-the-run bond (higher yield, cheaper price)
  • Sell on-the-run bond (lower yield, expensive price)
  • Weight to be duration-neutral
  • Wait for spread to narrow

Worked Example: 10Y On-the-Run vs. Off-the-Run Trade

Market Setup:

  • On-the-run 10Y JGB #362: Yield 0.80%, Price 100.00, DV01 = ¥9,200 per ¥1bn
  • Off-the-run 10Y JGB #361 (issued 3 months ago): Yield 0.83%, Price 99.70, DV01 = ¥9,150 per ¥1bn
  • Spread: 3bp (on-the-run premium)

Historical Analysis: The 3bp spread is at the wide end of the 1-3bp historical range, suggesting mean reversion opportunity.

Trade Construction:

  • Buy ¥10 billion off-the-run JGB #361 (cheaper)
  • Sell ¥10.05 billion on-the-run JGB #362 (expensive)

The slight size difference (10.05bn vs 10bn) makes the position duration-neutral:

  • Long DV01: ¥10bn × ¥9,150 = ¥91.5 million
  • Short DV01: ¥10.05bn × ¥9,200 = ¥92.46 million (approximately matched)

Expected Profit (spread narrows to 1.5bp):

If the spread tightens from 3bp to 1.5bp (1.5bp convergence):

  • P&L ≈ ¥91.5 million × 1.5bp = ¥137 million profit

Time horizon: Typically 1-3 months until next auction or market rebalancing.

Risks:

  • Spread could widen further if market stress increases liquidity premium
  • Off-the-run bond may be harder to finance in repo (potentially higher repo rate)
  • When on-the-run bond itself goes off-the-run (at next auction), you may need to unwind

3. Repo Arbitrage: GC vs. Special (SC) Spread Trading

Repo arbitrage exploits differences in financing rates between:

  • General Collateral (GC) repo: Standardized rate for generic JGB collateral
  • Special repo (SC): Lower (sometimes negative) rates for specific bonds in high demand

For detailed repo mechanics, see Chapter 2.10.

Understanding GC vs. SC Spreads

The SC spread is defined as:

\[\text{SC Spread} = \text{GC Rate} - \text{SC Rate}\]

A positive SC spread means the specific bond is “trading special” (tight supply → high demand to borrow).

Historical GC vs. SC Spreads (10Y JGB CTD, Illustrative)
Period GC Rate SC Rate (CTD) SC Spread Market Context
Jan 2020 (Normal) -0.05% -0.10% +5bp Mild specialness, sufficient CTD supply
Mar 2020 (COVID Crisis) +0.10% -0.50% +60bp Extreme CTD scarcity, market hoarding bonds
Jun 2021 (YCC Era) -0.10% -0.15% +5bp BOJ SLF prevented extreme specials
Dec 2022 (YCC Widening) -0.08% -0.35% +27bp Futures delivery month, CTD squeezed
Jun 2024 (Post-YCC) +0.12% +0.05% +7bp Normal market functioning restored
What Drives SC Spreads?
  1. Futures Delivery Cycles:
    • SC spreads widen dramatically in the 2-4 weeks before JGB futures delivery dates (March, June, September, December)
    • Short futures sellers scramble to borrow the CTD bond to deliver, pushing its SC rate deeply negative
  2. BOJ Holdings:
    • When the BOJ owns a large percentage of a specific issue (e.g., 40-50% during YCC peak), that bond becomes scarce
    • Even with the BOJ’s Securities Lending Facility (SLF), sustained demand can keep SC spreads elevated
  3. Short-Selling Activity:
    • If hedge funds build large short positions in a specific bond (betting on rising yields), they must borrow that bond in the repo market
    • High short interest → high SC spread
  4. Auction Cycles:
    • Immediately after a new 10Y JGB auction, the on-the-run bond often trades “special” because it’s the most liquid issue
    • The SC spread can reach 10-20bp for 1-2 weeks post-auction
Trading Opportunities from SC Spreads

Strategy 1: “Lending Out Your Bonds Special”

If you own the CTD bond and can lend it in the SC repo market at -0.30% while borrowing cash in the GC market at -0.10%, you earn a spread of +20bp risk-free.

Example:

  • Own ¥10 billion of JGB #362 (CTD)
  • Lend it via SC repo at -0.30% (you receive the bond back + pay 0.30% to borrower)
  • Borrow ¥10 billion via GC repo at -0.10% (you receive cash + pay 0.10% to lender)
  • Net spread: 20bp annualized

For 30 days: \(\text{Profit} = ¥10bn \times 0.20\% \times \frac{30}{365} = ¥1,643,836\)

This is essentially free money for bond holders during delivery squeeze periods.

Strategy 2: Basis Trade Enhancement

The SC spread directly affects the profitability of cash-futures basis trades (see Section 4.5). If the CTD bond goes “on special,” the cost to finance the position drops (or becomes a benefit), improving basis trade returns:

\[\text{Net Basis P&L} = \text{Gross Basis} + \text{Carry} - \text{Financing Cost}\]

When SC rate < GC rate, the financing cost component becomes positive (you earn from lending the bond), boosting overall returns.

Example:

  • Gross basis: ¥1.20 per ¥100
  • Carry (coupon - GC repo): ¥0.15
  • SC benefit (bond on special): ¥0.25
  • Total P&L: ¥1.60 per ¥100 (33% boost from SC specialness!)
Monitoring SC Spreads for Trading Signals

Professional JGB traders monitor SC spreads daily:

SC Spread Market Condition Trading Implication
< 5bp Normal market CTD readily available, no special opportunities
10-20bp Moderate tightness Delivery month approaching or high short interest; consider lending strategy
> 30bp Severe tightness Potential delivery squeeze; BOJ SLF may intervene; strong lending opportunity
> 50bp Crisis conditions Settlement fails likely; avoid short positions in that bond; extreme lending profits
Fails Risk and Strategic Failing

When SC spreads exceed -1.00% (extremely negative), it signals severe bond scarcity. Settlement fails become likely, creating operational risk for short-sellers who cannot source the bond even at punitive rates.

“Failing to the BOJ” Strategy:

In extreme scarcity situations, traders face a choice: pay extreme special repo rates to borrow the bond, or simply fail the delivery.

Break-Even Analysis:

If the SC repo rate is -2.00% (extremely negative) and the fail cost is GC + 150bp = 1.40%, it’s economically rational to fail intentionally rather than borrow at -2.00%:

$$\text{Cost to borrow in SC repo: } -2.00\% \text{ (you pay 2\% to borrow)}$$ $$\text{Cost to fail: } +1.40\% \text{ (you pay 1.4\% penalty)}$$ $$\text{Savings from failing: } 0.60\% \text{ (60bp)}$$

Historical Context:

This term arose during YCC when the BOJ owned so much of certain issues that fails became inevitable. Traders would strategically fail rather than pay punitive special repo rates, knowing the BOJ (as the largest holder) would be the one not receiving delivery. The BOJ eventually increased SLF operations to reduce this behavior.

BOJ’s Securities Lending Facility (SLF) as Circuit Breaker:

The BOJ’s SLF acts as a “release valve” for extreme specials:

  • Market participants can borrow specific JGBs from the BOJ at a fixed fee (typically 10-20bp)
  • When SC spreads widen beyond the SLF fee, arbitrageurs borrow from the BOJ and lend to the market, capping the spread
  • During March 2020 COVID crisis, SC spreads spiked to +60bp; the BOJ conducted emergency SLF operations, lending ¥2.5 trillion in specific issues within one week
Repo Arbitrage Execution

Classic Trade:

  1. Identify bond trading special (SC rate « GC rate)
  2. Borrow the bond via reverse repo at SC rate
  3. Lend it out via repo at GC rate
  4. Pocket the spread with minimal risk

Required Infrastructure:

  • Multiple repo counterparty relationships
  • Access to special repo market (not all participants have this)
  • Real-time monitoring of GC/SC spreads
  • Operational capacity to manage fails and SLF borrowing

This is low-risk income from rate differentials, though availability of special bonds can be limited to established market participants.

4. Cross-Market Arbitrage

Cross-market arbitrage involves exploiting price differences between related markets:

A. JGB vs. Interest Rate Swaps

JGB yields and JPY interest rate swap (IRS) rates should maintain relatively stable spreads. When anomalies occur:

Trade:

  • Buy JGB (if swap spread is too wide)
  • Pay fixed on IRS (receive floating)
  • Profit when spread normalizes

Covered in detail in Chapter 7.2 and 7.5.

B. JGB vs. Other Sovereign Bonds

Less common, but currency-hedged arbitrage between JGBs and other sovereigns (U.S. Treasuries, German Bunds) can exist when:

  • Credit spreads diverge from fundamentals
  • Supply/demand imbalances create distortions
  • Central bank policy divergence creates temporary mispricings

Challenges:

  • Currency hedging costs (FX swaps)
  • Different settlement conventions
  • Time zone and liquidity differences

Why True Arbitrage is Rare in JGB Markets

The JGB market is highly efficient due to:

  1. Sophisticated participants: Major banks, dealers, and hedge funds continuously scan for mispricings
  2. Low transaction costs: Tight bid-offer spreads, especially in on-the-run issues
  3. Deep liquidity: Large volumes traded daily, allowing quick execution
  4. Technology: Algorithmic trading systems identify and exploit tiny discrepancies in milliseconds
  5. BOJ presence: Central bank operations keep certain relationships stable (especially during YCC period)

When “arbitrage” opportunities appear, they typically:

  • Involve execution risk (prices move before all legs complete)
  • Require capital (not truly zero investment)
  • Carry model risk (assumptions about relationships may break)
  • Have limited size (large positions move prices against you)

Relative Value vs. Pure Arbitrage

Most JGB “arbitrage” strategies are actually relative value trades:

Feature Pure Arbitrage Relative Value
Risk Zero (theoretically) Low but non-zero
Capital Required Zero (self-financing) Positive investment needed
Holding Period Instant to delivery Days to months
Profit Source Mispricing correction Mean reversion, spread convergence
Example CTD delivery arbitrage On-the-run / off-the-run

Understanding this distinction is crucial for risk management and return expectations.

Execution Challenges in Arbitrage Trading

Even when apparent arbitrage opportunities exist, execution challenges can erode or eliminate profits:

1. Bid-Offer Spreads

Each leg of a multi-leg arbitrage incurs bid-offer costs:

Example:

  • On-the-run 10Y: 2 tick spread
  • Off-the-run 10Y: 3 tick spread
  • Total cost: 5 ticks = 0.05 price points

For a ¥10 billion trade, this is ¥5 million in transaction costs, which must be recovered from the spread convergence.

2. Leg Risk

Executing multiple legs simultaneously is difficult:

  • First leg executes, market moves before second leg completes
  • You’re left with unhedged directional exposure
  • Must decide whether to complete at worse prices or abandon the trade

Professional traders use:

  • Algorithmic execution: Automated systems execute all legs simultaneously
  • Futures for one leg: More liquid, easier to execute quickly
  • Prime broker relationships: Better pricing and execution priority

3. Financing Costs

“Arbitrage” positions often require funding:

  • Repo rates may be higher than expected
  • Haircuts require additional capital
  • Margin requirements for futures positions

These costs must be factored into return calculations.

4. Model Risk

Arbitrage strategies rely on assumptions:

  • Historical spreads will mean-revert
  • Correlation relationships remain stable
  • CTD won’t switch unexpectedly

During BOJ policy transitions (e.g., YCC exit in 2024), many historical relationships broke down temporarily, causing losses for arbitrage strategies.

Practical Implementation

For those implementing arbitrage strategies:

  1. Start small: Test strategies with limited capital before scaling
  2. Monitor constantly: Spreads can move quickly; set alerts and stop-losses
  3. Understand all costs: Include bid-offer, repo, commissions, margin
  4. Have multiple exit plans: Don’t rely solely on convergence
  5. Track hit rates: Measure how often trades work vs. fail
  6. Use technology: Manual execution is too slow for most opportunities
  7. Know your counterparties: Repo and borrowing relationships matter

Key Takeaways

  1. True arbitrage is rare in efficient JGB markets; most opportunities are relative value trades
  2. On-the-run/off-the-run spreads offer the most accessible relative value opportunities
  3. Repo arbitrage requires deep understanding of special vs. GC dynamics
  4. Execution is critical: Transaction costs and leg risk can eliminate small edge
  5. BOJ policy matters: Central bank actions can disrupt traditional arbitrage relationships

Arbitrage and relative value strategies require significant infrastructure, market access, and expertise—but when executed well, they provide steady, low-volatility returns that complement directional strategies.

Chapter 4 Summary

We’ve covered five major JGB trading strategies:

Each strategy has distinct risk-return characteristics and works best in specific market environments. Professional JGB portfolio managers combine these strategies to generate returns while managing overall portfolio risk.

Next Chapter

Chapter 5 - Economic Indicators →