Why Convertible Bonds?
In the Chinese A-share market, three asset classes let retail investors sleep at night: bank wealth-management products, money-market funds, and convertible bonds. The first two pay almost nothing. The third hands you something rare — a guaranteed floor with uncapped upside. That is exactly what has made it the most underrated, yet most retail-friendly, asset of the past decade.
A convertible bond is not a get-rich-quick scheme. Its real edge is structural: losing is hard, winning is not, and T+0 keeps it simple. This book hands you a discipline: buy low, spread wide, turn over rarely — turning a niche instrument into a repeatable income stream.
- Rule 1: Every example, rule and number refers to mainland China A-share convertible bonds (Shanghai & Shenzhen). B-shares and offshore CBs are out of scope.
- Rule 2: Red = up, green = down. This is the Chinese convention and the reverse of the West. Remember it.
- Rule 3: Prices are in renminbi (¥). One lot = 10 bonds = ¥1,000 of face value.
What this book promises
- ✅ Every public rule of the convertible bond market, explained plainly.
- ✅ A low-risk strategy stack validated across a decade of live data.
- ✅ At least 10 traps that look attractive and are in fact lethal.
- ✅ A final psychology checklist to protect the money you make.
What Is a Convertible Bond?
1.1 One-sentence definition
A convertible bond ("CB") is a bond issued by a listed company that can, at some future date, be converted into that company's shares at a pre-agreed price. It carries three identities at once:
1.2 One chart: why a convertible bond has a floor
1.3 Why it is special in China
- T+0 same-day round trip — buy and sell within the session, avoiding T+1 overnight risk.
- No daily price limit (except listing day) — the ceiling tracks the underlying stock.
- ~0.01% commission (both sides) plus zero stamp duty — very cheap to trade.
- Credit-based subscription — since 2017 retail investors can subscribe at maximum quota with no shareholding requirement: near-zero barrier to entry.
- The downward-revision clause — the issuer may cut the conversion price, effectively handing retail holders a free escape hatch. A uniquely A-share gift.
The Nine Core Features
After this chapter, you will read any convertible bond's data page on Jisilu, Eastmoney or Ningwen at a glance.
2.1 Three "souls": floor, equity, option
- Bond floor (pure bond value) — the present value of all future cash flows if treated as an ordinary bond. As price approaches the floor, it stops falling.
- Equity value (conversion value) — what you would get by converting immediately and selling the shares.Conversion value = (100 / conversion price) × share price
- Option value (time value) — the chance that the stock rises before you convert.Implied option value ≈ price − conversion value
2.2 The six vital signs
| Dimension | Feature | Plain meaning | Ideal range |
|---|---|---|---|
| Face value | ¥100 | 1 lot = ¥1,000 face; the smallest tradable unit | Fixed |
| Coupon | 0.2%–0.6% in year 1 | Annual interest received (taxed at 20%) | Step-up schedule |
| Outstanding size | ¥0.1–3bn | How much is left in the market. Tiny = wild, huge = dull | ¥300m–1.5bn |
| Remaining life | 1–6 years | The 1–2 year window is the most explosive for call games | 1.5–4 years |
| Conversion premium | formula | (price / conversion value − 1) × 100% | < 30% for low-buy |
| Call / put / maturity redemption | Read together — see Chapter 6 | ||
2.3 How is the conversion premium computed — and why does it matter?
It is the single most important valuation gauge.
Conversion premium = (CB price / conversion value − 1) × 100%It tells you how much further the stock must rise before the bond's price catches up with its conversion value:
- Price 110, conversion value 100 → premium 10% → the stock needs +10% for the bond to reach parity.
- Price 120, conversion value 100 → premium 20% → the bond now tracks the stock almost 1:1; the option value is spent.
- Price 130, conversion value 100 → premium 30% → you are in call-game territory (Chapter 6).
Seeing "price ¥98, looks cheap" and piling in is looking at price alone. Always read the conversion premium in the same glance: if the stock has collapsed and the premium is 60%, then ¥98 is actually expensive — the stock would have to rally 60% just to break you even.
Trading Rules & T+0 in Practice
3.1 Shanghai vs Shenzhen: three small differences
| Item | Shanghai (11/113/118) | Shenzhen (12/127/128) |
|---|---|---|
| Trading unit | 1 lot (10 bonds) | 1 lot (10 bonds) |
| Order type | Limit price + quantity | Limit price + quantity |
| Opening auction (listing day) | 9:15–9:25, cap ¥150 | 9:15–9:25, cap ¥130 |
| Listing-day band | −43.3% to +57.3% | −43.3% to +57.3% |
| Subsequent days | ±20% (halts to ±30%) | ±20% |
| Conversion code | Derived from the original code (190×××) | Separate conversion code |
3.2 The three-phase opening auction
- 9:15–9:20 — orders may be placed and cancelled. Everything you see is phantom liquidity; do not trust it.
- 9:20–9:25 — orders may be placed but not cancelled. This is real intent.
- 9:25 — the opening price prints. No orders between 9:25 and 9:30.
3.3 Circuit breaker (applies on both exchanges)
Touch ±20% intraday → 30-minute halt. Touch ±30% → halted until 14:57. After 14:57 no further halts apply and trading is free.
- After winning a new issue, place your sell order around 9:24:30, probing at roughly "previous close × 1.2".
- Day traders should play only the final window between the two halts — never chase the open.
- Low-buy investors: be patient with limit orders. Never use market orders.
The IPO Subscription Strategy
Since credit-based subscription arrived in 2017, convertible bond IPOs have been almost barrier-free: no shareholding requirement, no cash frozen, pay only if you win. "Subscribe for ¥1m of face value, win one lot" became free money for ordinary people.
4.1 The subscription workflow
4.2 How much can you actually make?
Backtested data (Shanghai & Shenzhen, 2020–2023):
| Period | Win rate | Avg. lots per account / yr | Avg. first-day gain | Annual expected P&L |
|---|---|---|---|---|
| 2020 bull | ~ 0.003% | 15–30 | +25% | ¥3,000–6,000 |
| 2021 choppy | ~ 0.005% | 40–80 | +22% | ¥5,000–12,000 |
| 2022 credit squeeze | ~ 0.008% | 60–120 | +18% | ¥4,000–10,000 |
| 2023 steady | ~ 0.006% | 50–100 | +20% | ¥5,000–12,000 |
Note: a single account earns roughly 1.5%–4% a year from subscriptions — crumbs, but risk-free crumbs. Treat it as a base layer, never as your main engine.
- Three brokerage accounts under one ID number = three independent quotas.
- You and your spouse = two quotas.
- Subscribing for ¥1m of face value vs ¥100k makes almost no difference to your win rate — so always go max.
- Allotment numbers appear after 18:00 on T+2. Check your broker app or SMS.
The Low-Price Trio
These are the three metrics every medium-to-long-term, low-risk convertible bond investor must master. Any "exotic bond" gambling that ignores them is outside the system in this book.
5.1 Sword one: price ≤ ¥100
When the price falls to about ¥100, the market has priced in a bear-market panic:
- Floor protection — the pure bond value catches further falls.
- Optionality — the downward-revision clause gives the issuer an incentive to push the stock up and cut the conversion price, delivering a 30%+ recovery.
- Historical evidence — in the two sub-par collapses of 2018 and 2022, average two-year returns were 25%–40%.
5.2 Sword two: the "double-low" score
Double-low = price + conversion premium (percentage points)It fuses the bond and equity dimensions into one number:
- Double-low < 110 — undervalued, buy candidate.
- 110–120 — fair, may build a position.
- 120–130 — pricey, small allocation only.
- > 130 — call-game zone; you are a seller, not a buyer.
5.3 Sword three: yield-to-maturity > 0
Discount the maturity redemption price plus all remaining coupons back to today. If the resulting YTM > 0, you are holding what amounts to a principal-protected deposit paying 1%–3%.
The Four Core Games
This is the most seductive part of the convertible bond market. Understand these four mechanisms and you can explain 90% of all wild price moves.
6.1 Mandatory redemption — the main profit event
Trigger: the stock closes at or above 130% of the prevailing conversion price on at least 15 of any 30 consecutive trading days.
- On trigger, the issuer may force-redeem the bonds at face value plus accrued interest (usually ¥100.5–103). You must convert or sell before the deadline.
- Practical implication: the ¥130–200 band is the "call-game zone". Once the call is announced and the stock breaks back below the trigger, the bond can collapse towards ¥130 or even ¥100.
6.2 Downward revision — the "free money" clause
Trigger: usually the stock closes below 85% of the prevailing conversion price on at least 15 of any 30 consecutive trading days; the board may then propose cutting the conversion price.
- Issuers want conversion (it avoids repaying cash at maturity), so revisions are often pushed through.
- Classic play: sit in a bond whose stock is under pressure and wait for the revision announcement — the bond often jumps +10% to +30% on the day.
6.3 Putback — the exit in the final two years
Trigger: during the last two interest periods, if the stock closes below 70% of the conversion price for 30 consecutive trading days, holders may put the bonds back to the issuer at face value plus accrued interest.
The put price is typically ¥101–103. This floor gives investors a backstop and pushes issuers to revise or to support the stock.
6.4 Default — low probability, black swan impact
Landun, Soute and Hongda are the historical cases where convertible bonds defaulted outright. When the underlying stock is delisted, the bond loses its repayment capacity. See the rating + size + sector three-gate filter in Chapter 9.
After the issuer announces it "will exercise the redemption right", typically only 5–10 trading days remain.
- Already converted: sell the shares.
- Still holding bonds: sell immediately after day one, even if it means leaving a little on the table.
- Greedy holders waiting for ¥200 usually round-trip to ¥130, or worse, ¥110.
Arbitrage Strategies
Once you are fluent in subscriptions, low-buy and call games, arbitrage is how you turn an informational edge into cash.
7.1 Conversion-discount arbitrage (only while halted)
Process:
Risk: an overnight gap on T+1. If the discount is small and the stock opens lower, the trade can lose more than the commission saved.
7.2 Rights-allocation arbitrage
Holding the stock on the record date entitles you to a pro-rata allocation of the new convertible bond. In theory you buy the stock, receive the bond, and sell the bond the next day. Beware:
- The stock typically gaps up into the record date and down afterwards — you can get trapped.
- You need to model the stock's valuation, the issue size and the allocation ratio precisely. Experienced hands only.
7.3 The "last-days" play in a call game
Between the call announcement and the record date, the bond usually trades pinned to its conversion value. If the stock moves sharply intraday, a T+0 arb is still possible — but always with a stop-loss in place.
- 99% of the opportunities are consumed by machines. The window left for retail is seconds to minutes.
- Arbitrage is a speed and infrastructure game, which is not a retail edge. Treat it as a side dish, never an ALL IN.
The Risk Map
Every risk, written out in three parts: trigger + typical outcome + how to avoid it.
Outcome: hold above ¥130 and you are redeemed at ¥100.
Fix: take profit in tranches from ¥122.
Outcome: the bond falls to ¥30–50 and stops.
Fix: only AA- or better; never an ST issuer.
Outcome: the game fails, limited price repair.
Fix: check the revised conversion value and whether it still trades below par.
Outcome: technically convertible, but nobody bids.
Fix: avoid micro-cap, poor-quality issuers.
Outcome: stock −5%, bond −20%.
Fix: no purchases above a 30% premium.
Outcome: you cannot get filled either way.
Fix: prefer ¥300m–3bn issues.
Eight Hard Screening Criteria
Chapters 1–8 gave you the rules and the strategies. This chapter is the sieve. Only names passing all eight criteria enter the portfolio.
| # | Criterion | Threshold | Why |
|---|---|---|---|
| 1 | Price | ≤ ¥115 | Keep equity value while staying near the floor |
| 2 | Conversion premium | ≤ 30% | Do not overpay |
| 3 | Credit rating | ≥ AA- | Avoid credit blow-ups |
| 4 | Outstanding size | ¥300m–3bn | Balance liquidity and beta |
| 5 | Sector | Not sunset / not policy-exposed | Still alive in five years |
| 6 | Underlying ROE | ≥ 5% | The company is not losing money |
| 7 | Yield to maturity | YTM ≥ 0 | Holding the bond alone pays |
| 8 | No ST flag, no heavy selling pressure | Yes | Avoid forced landmines |
- Open the convertible bond screener on Jisilu (jisilu.cn) or Eastmoney (data.eastmoney.com).
- Go to the "Convertible Bonds" page → "Screen".
- Enter all eight thresholds above; you will be left with 5–20 names.
- Manually remove any with "major shareholder selling in the past month" or "unlock within three months".
- The remaining 5–10 names become your watchlist — wait for a pullback and scale in.
Position Management for High Win Rates
The same strategy, sized differently, produces wildly different win rates. This chapter is the last line of defence for the profits you make.
10.1 Three iron rules
10.2 Double-low rotation
Once a week, sell the holdings whose double-low score has lost its edge and buy the lowest-scoring newcomers.
- ¥100k of capital → spread across 10 names, ¥10k each.
- Rebalance once every Sunday; total trading cost stays below 1%.
- Historical annualised return: 12%–18% excluding subscriptions.
10.3 Grid trading example (ideal in a ¥110–125 range)
Grid: ceiling ¥125, floor ¥105; midpoint ¥115; step ¥2.50; ¥10k per grid- Every ¥2.50 drop: add ¥10k.
- Every ¥2.50 rise: trim ¥10k.
- If the price breaks four grids in either direction, pause the grid and reassess whether a call game has begun.
Six Real Case Studies
Six stories covering the three archetypal endings: winners, disasters, and games played close to the line.
What happened: the stock ground lower for six months, then the company announced the board had proposed a downward revision of the conversion price. The bond jumped 12% the next day, and as the stock rebounded the bond entered the call zone within four months.
Lesson: a revision play really can double your money — provided (a) the underlying fundamentals are intact, (b) the issue is small, and (c) the credit rating is safe.
What happened: EV order momentum drove the stock up 60%, and the bond entered the call zone in tandem. The classic retail error is refusing to sell above ¥200 — the bond then round-tripped hard after the announcement.
Lesson: once the redemption is announced, be flat within five days.
What happened: the issuer posted consecutive losses, a qualified audit opinion and a cascade of debt defaults. The stock was delisted in 2023 and the bond defaulted, with holders recovering about 5 cents on the yuan.
Lesson: a B rating, two straight loss-making years, irregular guarantees — never touch it. An AA- rating is a floor, not a target.
What happened: the stock was stringing together limit-ups and the bond spiked to ¥150. But Soute was rated BB, was loss-making, and carried irregular guarantees. When the streak ended the bond stepped down, and eventually defaulted on delisting.
Lesson: you may gamble on an "exotic bond", but never with size, and never overnight.
What happened: buyers piled in on resumption, but the announcement revealed five qualifying days above 130% had already accrued. The stock pulled back and the bond collapsed.
Lesson: do not build positions around ¥130. Either buy early and wait for the game, or wait for the announcement and exit. The middle stretch is where retail goes to die.
What happened: China Merchants Bank issued a ¥15bn convertible with a rock-solid underlying. The play was simple: buy low, let the stock grind up, collect coupons, be patient. No call game, pure bond-plus-slow-bull.
Lesson: not every convertible bond needs a call game. Good paper deserves a slow dance.
Myths & Trading Psychology
12.1 Ten fatal mistakes
- Trading it like a stock — forgetting the bond floor exists.
- Ignoring the conversion premium — buying only because the price "looks cheap".
- All-in on one name — treating a revision play as a lottery ticket.
- Refusing to sell — still holding at ¥200, ¥300.
- Chasing exotic micro-caps — anything under ¥50m outstanding is a casino.
- Never checking the rating — buying a B-rated bond is gambling, not investing.
- Rights-arb without modelling — letting a gap-down in the stock eat your profit.
- Acting on tips — "my friend says the revision is coming" is not a thesis.
- No stop-loss — holding and hoping until the position cannot recover.
- Using margin — T+0 convenience plus forced liquidation is a fatal combination.
12.2 Five psychological rules
12.3 Three sentences worth carving in stone
1. Never buy a ¥100 bond for the long term unless a downward revision is genuinely plausible.
2. A call is not profit — it is an umbrella. Close it before the rain.
3. A real master sells at ¥200.
Glossary / Tools / Data
A. Key terms
| Term | Meaning |
|---|---|
| Conversion price | The share price at which one bond converts. |
| Conversion value | = 100 / conversion price × share price. |
| Conversion premium | = price / conversion value − 1. |
| Bond floor | The present value of all future cash flows, treated as a straight bond. |
| YTM | Yield to maturity — higher is better. |
| Mandatory redemption | The issuer's right to force redemption at par, demanding conversion. |
| Downward revision | The issuer lowers the conversion price, lifting conversion value. |
| Putback | Holders may sell back at par plus accrued interest once triggered. |
| Maturity redemption | The issuer redeems at the maturity redemption price. |
| Circuit breaker | The ±20% / ±30% trading-halt mechanism. |
B. Tool kit
- Jisilu (jisilu.cn) — the most complete convertible bond database.
- Eastmoney (data.eastmoney.com) — live quotes.
- Ningwen (ningwenjing.com) — historical prints and the redemption calendar.
- Broker apps — keep at least two independent accounts (max-quota subscription and conversion speed).
- ChinaBond (chinabond.com.cn) — ratings and issuer fundamentals.
C. Recommended three-month roadmap
D. Acknowledgements
Sources include public material from Jisilu, Ningwen and Eastmoney communities, plus bank convertible bond offering documents and company announcements. All figures are illustrative; backtest data is based on public samples. Investing carries risk — make your own decisions.
Reading is not earning — go and do it
The real finish line of this book is not the twelve chapters and the appendix. It is the moment you open your brokerage account today and submit your first subscription.
Convertible bonds will not make you rich overnight. They can give you one steady thing 80% of retail investors lack: a sense of safety. Go slow to go fast.
Step 1: subscribe to next Thursday's new issue — at maximum quota.
Step 2: run the 8 criteria on Jisilu and shortlist 10 names.
Step 3: wait for a pullback and scale in.