Trading Principles: Rationale, Stops, and Why Consistency Beats Cleverness
Fixed Income Trading and Risk Management by Alexander Düring (ISBN 9781119756354)
Previous: Swaps | Next: Curve Trading
The book shifts from instruments to process. Chapter 30 is Düring’s trading philosophy chapter, and it is refreshingly blunt about luck.
A trade is a voluntary bet on market risk you plan to close at a profit (or accept a loss). Not trading when new information hits is also a trade. P&L includes carry and cash along the way.
With enough traders, someone will look like a genius by chance alone. Copying their strategy is not science. Still, some ingredients recur.
Four ingredients of a real trade
- Rationale: Why should this work? That is why the prior 27 chapters exist.
- Profit target and stop-loss: Know when you win and when you cut.
- Exit strategy: How you get out and what it costs, decided upfront.
- Consistency: Instruments and size must match the analysis.
“Let profits run, cut losses short” means trailing stops: as P&L rises, move the stop up so a reversal does not give it all back. Hit target? Move stop to target and aim higher, or close if the trade goes sideways and balance sheet is better elsewhere.
The “greater fool” strategy (assume someone else takes your bag) is excluded here. No analysis required.
Liquidity matters in the exit plan. A crisis hedge that needs rare bonds or wide multi-leg packages will pay frictional costs exactly when stress hits.
Consistency traps in fixed income
Finite-maturity instruments breed implementation bugs:
- Model error: Spline says sector X is cheap, but every bond in X looks rich on the spline. Maybe the model is wrong, not the bonds.
- Basis: You analyzed bonds but traded futures. Basis move can eat the bond view.
- Sample bias: View on AA banks, trade one AA bank bond. Idiosyncratic risk piled on.
- Carry: You sized for positive carry against the original spread weights. Trade no longer matches the thesis.
Finding trades
Stat arb: mine history for deviations from mean, bet on reversion. Easy to automate, crowded, competes with smart-beta algos. Also misses half the opportunity: you skip the initial move away from fair value.
Ad hoc: use non-price information (policy, supply, regulation). Hope human edge beats machines.
Common sense blend: dislocation plus a story. A wide spread might revert, or it might be the start of a regime change. Ask why before sizing.
Instrument choice: default to liquid benchmarks, but cheap alternatives can work if the thesis is relative value.
Trade portfolios
You will not bet the whole book on one idea. Trades compete for capital; some risks offset.
Black-Litterman and mean-variance help, but Düring flags epistemic problems: single-period optimizers vs multi-horizon trades, non-normal payoffs, and the fact that stat-arb and trend bets imply you should revise means and correlations, not just means.
Luck and sample size
Düring cites Taleb-style skepticism: streaks happen. A trader with sixty percent hit rate over two years might be skilled or might have sized small on losers and gotten lucky on one fat tail. The chapter does not tell you how to detect skill statistically. It tells you not to outsource conviction to someone else’s track record without understanding their stop discipline and book concentration.
Where this fits in the book arc
Parts I-IV built instruments. Part V (money market through securitized products) built balance sheet context. Part VI starts with process before more tickets. That ordering is intentional. Junior desks often want butterfly formulas; seniors want to know whether the butterfly matches the risk meeting story.
Chapter 30 is not a strategy manual. It is a checklist before you touch a ticket. The rest of Part VI (curve trades, bond RV, PCA, portfolio tools) assumes you already bought into this discipline. If you skip it, you can price a butterfly perfectly and still lose money because your hedge instrument did not match your story.