Explicit Rational Formulae for Bachelier (Normal) Implied Volatility

Fabien Le Floc’h · 2026

Read Explicit Rational Formulae for Bachelier (Normal) Implied Volatility by Fabien Le Floc’h (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-18343.txt.

Introduction

In a 2026 paper, Fabien Le Floc'h gives two direct formulas for Bachelier, or normal, implied volatility in option pricing. Each formula takes price, forward, strike, and expiry as inputs and returns volatility without iteration. One version stresses accuracy in the far tail, while the other splits the near-the-money branch for speed. Tests in double precision keep both close to machine accuracy.

Catalog record

CREATORFabien Le Floc’h
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE18520
LICENSEhttp://creativecommons.org/licenses/by-sa/4.0/
RIGHTSCC BY SA
SOURCEarXiv

Source: https://arxiv.org/abs/2605.18343

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Explicit Rational Formulae for Bachelier (Normal) Implied Volatility Fabien Le Floc’h Abstract We present two explicit rational formulae for Bachelier, or normal, implied volatility. The formulae take the option price, forward, strike, and expiry as inputs and return the implied normal volatility without iteration. They follow the branch structure of LFK-4, but use the simpler near-the-money variable given by the absolute forward-strike difference divided by the tail time value, avoiding a logarithm and a small-argument Taylor branch in that region. LFK-2026 is the accuracy-oriented formula and approximates reciprocal absolute standardized moneyness directly in the far tail. LFK-2026C keeps the same shifted out-of-the-money rational tail approximation, but splits the near-the-money branch into a very small low- u u rational and a mid-range rational. In double precision tests both remain close to machine accuracy, while LFK-2026C is the faster scalar implementation on the current benchmark mix. keywords Bachelier model; normal implied volatility; bp vol; implied volatility; rational approximation; computational finance † † firstpage: 1 † † volume: 1 † † issue: 1 † † articlenumber: 0 † † year: 2026 † † copyright-year: 2026 † † authornames: Fabien Le Floc’h † † address: Independent researcher; [email protected] † † corresponding: Correspondence: [email protected] † † reftitle: References \fancypagestyle plain \fancyhf \fancyfoot [C] 1 \fancyhead \fancyfoot \fancyfoot [C] 1 \newgeometry left=2.5cm, right=2.5cm, top=1.27cm, bottom=1.08cm, includehead, includefoot, footskip=36pt, headsep=24pt \fancyheadoffset 0pt \fancyfootoffset 0pt 1 Introduction The Bachelier model, introduced in Bachelier’s 1900 thesis bachelier1900 , is widely used when prices are naturally described by a normal volatility rather than a lognormal volatility. Its call price is explicit once the volatility is known, but the inverse problem—recovering the normal volatility from an observed option price—has no elementary closed form. A root finder is robust, but it is often too expensive for calibration, risk, and large grid calculations where the inverse must be evaluated millions of times. This is not only a numerical issue. In interest-rate derivatives, normal or Bachelier volatilities are the usual convention when forwards, strikes, or rates can be close to zero or negative. Empirical work on low-rate swaption markets therefore considers Black, Bachelier, and shifted-lognormal volatility quotes across currencies including EUR, USD, GBP, and JPY patel2018quotes ; in particular, the convention has long been familiar in Japanese-yen interest-rate markets. The same modelling issue briefly appeared outside rates in April 2020, when WTI crude oil futures traded below zero. CME Clearing then announced a switch of its options pricing and valuation model to Bachelier to accommodate negative underlying futures prices and negative strikes cme2020bachelier ; Choi, Kwak, Tee, and Wang choi2022guide review this temporary commodity-exchange transition from Black–Scholes to Bachelier. In interest-rate markets the same implied quantity is often quoted as basis-point volatility , or bp vol. This is not a separate volatility model: it is the normal, equivalently Bachelier, volatility expressed in the natural absolute units of the underlying. For a rate forward quoted as a decimal, a normal volatility \sigma corresponds to 10^{4}\sigma basis points per square-root year. This contrasts with Black or lognormal volatility, which is dimensionless and measures proportional rather than absolute moves. We use the standard option terminology throughout: at the money (ATM) means F=K ; for a call, in the money (ITM) means F>K , and out of the money (OTM) means F<K . The formulae are written with the OTM time value, so the same expressions also cover the reflected ITM side through put–call symmetry. Explicit approximations fill this gap. The LFK-4 approximation lefloc2016basispoint is a standard reference: it transforms the inverse problem into a small number of rational approximations, using one branch near the money and three branches in the OTM tail. Related work includes the direct arithmetic Brownian-motion implied-volatility approximation of Choi, Kim, and Kwak choi2009 . That work is useful as a compact closed-form reference, but it is not aimed at the near-machine-epsilon accuracy sought by LFK-4 and by LFK-2026 here. For context, Jäckel’s rational inversion of the Black–Scholes formula jaeckel2017 shows the same high-accuracy philosophy in a different model; his normal implied-volatility approximation is also a popular production reference. QuantLib quantlib includes an implementation in this family, but it is not a fully faithful transcription of Jäckel’s original code and is less accurate. The timings below therefore compare the local Java implementations used in this study rather than the QuantLib port. The goal here is to keep the practical LFK-4 structure while improving the accuracy–cost trade-off. The two formulae below are designed to be portable: they use only arithmetic operations and a logarithm in the OTM routing variable. Both LFK-2026 and LFK-2026C avoid the final OTM square root by fitting 1/|d| directly. Accurate validation of these formulae requires an accurate forward Bachelier price. Direct formulas for the OTM time value suffer cancellation in the tails, which can make an inverse approximation look worse than it is. For benchmarking we therefore compute the OTM time value with a symmetric \operatorname{erfcx} expression; this makes the comparison primarily about the inverse formula rather than about loss of precision in the price used as input. 2 Setup and Notation Let x=F-K,\qquad v=\sigma\sqrt{T},\qquad d=\frac{x}{v}. The undiscounted Bachelier call price is C=x\Phi(d)+v\phi(d). Thus ATM corresponds to m=|x|=0 . Away from ATM, the call is ITM for x>0 and OTM for x<0 ; the reflected put side gives the same time value. As in LFK-4, the inverse is expressed in terms of the out-of-the-money time value and the absolute moneyness, c_{\mathrm{otm}}=\begin{cases}C-x,&x>0,\\ C,&x\leq 0,\end{cases}\qquad m=|x|. At the money, where m=0 , the inverse is exact: \sigma=\frac{C\sqrt{2\pi}}{\sqrt{T}}. For m>0 , define g=\frac{c_{\mathrm{otm}}}{m},\qquad\alpha=0.15,\qquad\beta_{s}=-\ln\alpha,\qquad\beta_{e}=300\ln(10). The ITM/near-ATM branch is used when g>\alpha . The OTM branch uses \tilde{\eta}=-\frac{\ln g+\beta_{s}}{\beta_{e}-\beta_{s}}. The three OTM zones are \begin{array}[]{c|c}j&\tilde{\eta}\text{ range}\\ \hline\cr 1&[0,0.011)\\ 2&[0.011,0.105)\\ 3&[0.105,1].\end{array} For all coefficient tables, rational functions are written in ascending powers: R_{p,q}(y;a,b)=\frac{P_{p}(y)}{Q_{q}(y)},\qquad P_{p}(y)=\sum_{i=0}^{p}a_{i}y^{i},\qquad Q_{q}(y)=\sum_{i=0}^{q}b_{i}y^{i},\quad b_{0}=1. For numerical evaluation, the same polynomials may be evaluated by ordinary Horner steps in descending powers. 3 Reference: LFK-4 LFK-4 lefloc2016basispoint uses z=\frac{m}{m+c_{\mathrm{otm}}},\qquad\eta=\frac{-z}{\ln(1-z)}. For small z , the original algorithm evaluates \eta by a Taylor expansion; otherwise it uses \operatorname{log1p}(-z) . This avoids cancellation in the near-ATM limit. For g>\alpha , LFK-4 returns \sigma=\frac{m+c_{\mathrm{otm}}}{\sqrt{T}}\,h_{\mathrm{LFK}}(\eta), with a P7/Q5 rational. For g\leq\alpha , it fits d^{2} with three P9/Q7 rationals in \tilde{\eta} , using the zones listed above. 4 LFK-2026 LFK-2026 uses the no-log ITM variable u=\frac{z}{1-z}=\frac{m}{c_{\mathrm{otm}}}. For g>\alpha , it uses a P10/Q9 rational: \sigma=\frac{m+c_{\mathrm{otm}}}{\sqrt{T}}\,R_{10,9}(u;a^{26}_{I},b^{26}_{I}). In the OTM branch, LFK-2026 approximates 1/|d| directly: W^{26}_{j}(\tilde{\eta})=R_{10,9}(\tilde{\eta};a^{26}_{j},b^{26}_{j}),\qquad\sigma=\frac{m}{\sqrt{T}}\,W^{26}_{j}(\tilde{\eta}). This removes the final square root in the OTM branch. The method uses more coefficients in the ITM branch than LFK-2026C, but the two methods share the same OTM rational approximation and the same direct OTM accuracy envelope. 5 LFK-2026C LFK-2026C uses the same no-log ITM variable u=m/c_{\mathrm{otm}} , but splits the ITM branch at u=0.20 . For small u it uses a compact P4/Q4 rational, and otherwise it uses a P9/Q8 rational: \sigma=\frac{m+c_{\mathrm{otm}}}{\sqrt{T}}\,\begin{cases}R_{4,4}(u;a^{26C}_{L},b^{26C}_{L}),&u<0.20,\\ R_{9,8}(u;a^{26C}_{I},b^{26C}_{I}),&u\geq 0.20.\end{cases} In the OTM branch it uses the same shifted-zone 1/|d| approximation as LFK-2026: W^{26C}_{j}(\tilde{\eta})=R_{10,9}(\tilde{\eta};a^{26}_{j},b^{26}_{j}),\qquad\sigma=\frac{m}{\sqrt{T}}\,W^{26C}_{j}(\tilde{\eta}). This is the fast variant. It keeps the OTM accuracy of LFK-2026 while reducing the average cost of the ITM/near-ATM branch. 6 Accuracy and Timing The validation uses a symmetric \operatorname{erfcx} expression for the OTM time value. With a=m/v , c_{\mathrm{otm}}=e^{-a^{2}/2}\left(\frac{v}{\sqrt{2\pi}}-\frac{m}{2}\operatorname{erfcx}\left(\frac{a}{\sqrt{2}}\right)\right), which avoids cancellation in both deep ITM

Read the complete text