Survivorship Bias in a Backtest: The Delisted-Fraction Smell Test

Survivorship bias is the term every quant knows and almost nobody makes actionable. Here is the one-number diagnostic I run before I trust any cross-sectional backtest: count the dead names as a share of your universe over a decade. If it is near zero, your data quietly deleted the losers and every result is biased upward.

By Rayan Azhari · 9 min read · Essays

Latest Essays

Essays

Diesel at 175: What Syria's Fuel Price Rise Shows, and What It Does Not

On 13 September 2026 diesel in Syria went from 125 to 175 new pounds a litre, and by Sunday afternoon roads were closed in four governorates. This piece sets out what the record establishes about that price and what it does not: a pricing committee with a list of factors and no formula, a world price the ministry cited but did not show, a stated need of 300,000 barrels a day of which a fifth is itemised, two import shares from one ministry on one day, and a population with nothing between it and the price. Five charts, every figure with its publisher, unit and date.

2026-09-14

Sharpe Ratio Limitations: Why I Promote on Calmar, Not Sharpe

A candidate in my Titan system posted a Sharpe near 1.4, cleared every statistical gate, and still traced a drawdown so deep and so long that I could never have held it through the trough. Here is why Sharpe is blind to the path and the tail, the survival-metric suite (Sortino, Calmar, CVaR, CDaR) that sees what it misses, and the one rule it bought me: promote on Calmar lift, not Sharpe lift. All figures are illustrative and sanitised.

2026-09-11 · 9 min

Article 14: Vertical Postcodes: How the UK Address System Handles Stacked Property

A series mining the PhD thesis on London and UK office buildings (Azhari, 2025). Key takeaway. The UK postcode is a flat two-dimensional locator carrying three-dimensional cities on its back, and UPRNs plus the AddressBase Premium hierarchy fill the gap that lets digital property services work vertically.

2026-09-08 · 13 min

Risk of Ruin Monte Carlo: Resample the Cause, Not the Effect

Most risk-of-ruin Monte Carlo resamples a strategy's realised P&L, which quietly bakes in the good luck you are trying to stress and understates the tail. Here is the correction that made my honest drawdown distribution far fatter than my first one, plus the relative gate a long-only sleeve actually needs. First person, British English, illustrative numbers only.

2026-09-04 · 9 min

Article 13: Why I Used Linear Regression Over Random Forest on 6,000 Buildings: A Modelling Walkthrough

A series mining the PhD thesis London and UK Office Buildings: Investigating Energy Use and Landlord-Tenant Influences (Azhari, 2025). Random Forest and a simple linear regression finished in a statistical dead heat on electricity EUI, so interpretability tipped the decision to the simpler model.

2026-09-01 · 11 min

Walk Forward Optimization Is Not Automatically Out of Sample

Everyone sells walk-forward validation as proof a strategy is out-of-sample. It is not. I ran my own walk-forward pipeline on pure random-walk noise and it still produced a healthy positive stitched Sharpe, because out-of-sample is a property of provenance, not partitioning. Here is the random-walk control that tells you whether your pipeline, or your edge, produced the number.

2026-08-28 · 9 min

Essays in your inbox