Options & Volatility
Option Volatility and Pricing: Advanced Trading Strategies and Techniques
Natenberg turns volatility into a number you can actually trade: one standard deviation of price change, per year, in percent. Everything else in the book, the Greeks, the spreads, the hedging, falls out of that one reframe. It is the clearest explanation of option risk in print, and it will not hand you a trade.
The one idea
Most people who buy an option are betting on direction. Natenberg's point is that direction is the smaller half of the bet. An option is a bet on speed. If a stock has to travel from 100 to 105 for your call to matter, what you need is not just up, it is up fast enough, before time runs out. Volatility is the number that measures speed, and Natenberg's whole book exists to turn that fuzzy word into arithmetic you can do at a desk.
His definition is exact and it never changes across 588 pages: volatility is a standard deviation, expressed as a percent, annualized. A stock at 100 with a 20 percent volatility means that one year out, there is roughly a two in three chance it lands between 80 and 120, a 19 in 20 chance between 60 and 140. That single sentence is the whole engine. From it he gets daily and weekly expected moves (divide annual volatility by 16 for a day, 7.2 for a week, because there are about 256 trading days and 52 weeks in a year and volatility scales by the square root of time). A $45 stock at 37 percent volatility should move about $1.04 on a normal day. If it has not moved that much in five days, either the market went quiet or your 37 percent is wrong.
The second half of the idea is that there are two volatilities and they are not the same animal. Realized volatility is what the stock actually did, measured from settlement to settlement. Implied volatility is what the option's price says the market expects, found by running the pricing model backwards until the theoretical value matches the market price. His example: a 105 call, stock at 98.50, three months out. Your 25 percent volatility says the option is worth 2.94. It trades at 3.60. Rather than call it 0.66 too expensive, Natenberg makes you say it is 3.5 volatility points too expensive, because 28.50 percent is the number that produces 3.60. That translation, from dollars into volatility points, is the habit the book is trying to install. Once you have it, every option in the chain is comparable and you can see which one is actually the good buy.
What it actually teaches
Natenberg never labels it a system, but the same six-step sequence runs from Chapter 6 through Chapter 13 and it is the spine of the book.
TURN VOLATILITY INTO A DAILY NUMBER
Volatility is quoted annually, like an interest rate. To use it, scale it down. Divide the annual volatility by 16 for a one-day standard deviation (16 is the square root of 256 trading days) and by 7.2 for one week (the square root of 52). A $45 stock at 37 percent volatility gives $45 x 0.37 / 16 = $1.04 a day and $2.31 a week. Now check reality against it. You should see a move bigger than one standard deviation about one day in three, bigger than two standard deviations about one day in twenty, which works out to roughly once a month. Natenberg's five-day example, where nothing exceeded $1.04 even once, is a signal the 37 percent input is too high (the actual answer was 27.8 percent).
PRICE THE OPTION IN VOLATILITY POINTS, NOT DOLLARS
Run the model forward with your volatility estimate to get a theoretical value. Then run it backwards on the market price to get implied volatility. Compare the two numbers, not the two prices. His worked case: 100 call trading 5.40 at 27.51 percent implied, 105 call trading 3.60 at 28.50 percent. In dollars the 100 call looks more expensive. In volatility terms it is a full point cheaper, and that is the one that matters. He also gives three rules for how sensitive things are: a change in volatility moves the at-the-money option most in total points, moves the out-of-the-money option most in percentage terms, and moves long-dated options more than short-dated ones. His July 2012 gold table proves it: raising volatility from 14 to 18 percent takes the 1600 call from 41.65 to 51.60 (+9.95, or 24 percent) while the 1800 call goes from 0.78 to 3.05, up 291 percent.
PICK THE SPREAD FAMILY FROM THE SIGN OF VEGA
This is the rule that decides what you trade. If implied volatility is below your estimate, options are cheap, so you want positive vega: long straddles and strangles, short butterflies and condors, ratio spreads buying more than you sell, long calendar spreads. If implied volatility is above your estimate, options are rich, so you want negative vega: short straddles and strangles, long butterflies and condors, ratio spreads selling more than you buy, short calendar spreads. He then splits the world into four quadrants by the signs of gamma and vega, because gamma says whether you want the underlying to move and vega says whether you want implied volatility to rise. Positive gamma with negative vega, for instance, wants a wild underlying and a calm option market at the same time.
READ THE POSITION THROUGH THE GREEKS
Delta is directional risk, and it has four readings he treats as interchangeable: rate of change, hedge ratio, equivalent underlying position (100 deltas equals one contract), and rough probability of finishing in the money. Gamma is deltas gained or lost per one-point move, so it is magnitude and speed risk. Theta is dollars lost per day, and it accelerates hard at the end: an at-the-money option might decay 0.03 a day with three months left, 0.06 with three weeks, 0.16 with three days. Vega is dollars per one percentage point of volatility. Rho he says openly is the least important and few individual traders should worry about it. The rule that matters: gamma and theta always carry opposite signs and their magnitudes track each other. Either movement helps you or time helps you. Never both.
FIND THE BREAKEVEN VOLATILITY AND SIZE ON IT
Theoretical edge alone is meaningless because you can manufacture more of it by trading bigger. So he equalizes edge across candidates and asks how far wrong you can be before the profit disappears. In his Chapter 13 example, three negative-vega spreads on an underlying at 48.40 with a 56-day expiry and an 18 percent forecast all show about 6.00 of edge: a short straddle, a ratio call spread, and a long put butterfly. Their breakeven volatilities are roughly 21 percent, 23 percent, and 21.5 percent. That gap is the margin for error, and it dictates position size. His own rule: a spread with a 2-point margin gets done 10 times; the same spread with a 7-point margin can be done 100 x 50. He also states flatly that straddles and strangles are the riskiest spreads there are, bought or sold, because they carry the biggest gamma and vega.
HEDGE DELTA TO ZERO AND KEEP ADJUSTING
A theoretical edge is only a claim until you collect it, and you collect it by staying delta neutral. Buy the underpriced option, take the offsetting position in the underlying, and every time the delta drifts, trade back to flat. That forces you to sell into strength and buy into weakness mechanically. His 10-week worked example (see the table) ends with a profit of 89.21 against a model prediction of 89.00, but the original hedge alone lost 422.50. All the money came from the adjustments. His caveat is the important part: adjustments do not raise your expected return, they only cut the variance around it. Adjust more often and you get a tighter, more model-like result at higher transaction cost. Adjust rarely and you have the same odds with much wilder swings.
What it looks like Monday morning
The first thing that changes is how you read an option chain. Instead of scanning premiums, you convert the annual implied volatility into a daily expected move and ask whether the underlying has actually been doing that. Divide by 16. Compare against the last twenty settlement-to-settlement changes. If the stock is printing half-a-percent days while the options are priced for one-and-a-half percent days, you now have a specific, quantified reason to be a seller, and you can say exactly how many volatility points of cushion you have before you are wrong.
The second thing is that you stop describing trades by their names and start describing them by their two signs. An iron condor stops being 'an income strategy' and becomes short gamma, short vega, which tells you precisely what kills it: a fast move, or a jump in implied volatility, or both. Before you enter, you find the volatility at which the position breaks even, and you size the position off that gap rather than off a percent-of-account rule. If the gap is two points, you do it small. If the gap is seven points and seven points is historically rare in that market, you do it big. And you learn the 40 percent rule for sanity checks: an at-the-forward option is worth roughly 40 percent of one standard deviation, so a 65-strike, three-month option at 18 percent volatility should run about 65 x 0.18 x 0.5 x 0.4, which is 2.33. You can now price an option in your head on a phone call.
| Component | Calculation | P&L |
|---|---|---|
| Original hedge, the options | 100 x (3.85 - 5.00) | -115.00 |
| Original hedge, the stock | 50 x (97.70 - 103.85) | -307.50 |
| Nine weekly delta adjustments | Forced to sell rallies and buy dips | +467.55 |
| Interest paid to carry the options | -500 x 6% x 70/365 | -5.75 |
| Interest earned on the short stock | +4,885 x 6% x 70/365 | +56.21 |
| Interest on the adjustment cash flows | Net across nine trades | -5.27 |
| Total, discounted back to today | 90.24 / (1 + 0.06 x 70/365) | +89.21 |
| What the model predicted | 100 x (5.89 - 5.00) | +89.00 |
The opening trade lost 422.50. Every dollar of profit came from the rehedging, and it landed within 0.21 of the model's prediction.
Volatility is just a trader's term for standard deviation.Sheldon Natenberg, Option Volatility and Pricing
Where it fails
- Without a statistics background, half this book is decoration. Natenberg says he uses an intuitive approach, and Chapters 6 through 13 mostly deliver on that. Then Chapter 18 gives you the full Black-Scholes formula with N(d1) and N(d2), Chapter 19 builds binomial trees, and Chapter 20 hands you an exponentially weighted moving average with a decay factor of 0.94 before pointing at GARCH and saying it is beyond the scope of the text. His own preface admits it: this book 'is in no way meant to take the place of a good university textbook on financial engineering.' If the phrase 'annualized standard deviation of logarithmic returns' is not already a sentence you can parse and compute, you will read Chapter 6, feel it click, and then hit a wall around page 340 that no amount of rereading fixes. The people this quietly does not work for are self-taught traders with no stats coursework and no spreadsheet habit, which is most of the retail audience the book is marketed to.
- The edge he describes is a market maker's edge, and he admits it in a footnote. The engine of the whole book is buying options below theoretical value and dynamically hedging to collect the difference. In Chapter 8 he rehedges 100 contracts every week for ten weeks and nets 89.21. Then, in a footnote, he writes that this ignores 'the very real advantage the professional trader often has from being able to buy at the bid price and sell at the ask price. A retail customer can never hope to match the profit resulting from this advantage, nor should he try to do so.' He is being honest, but it undercuts the premise. Natenberg spent his career as a floor market maker at the CBOE and later at Chicago Trading Company. The frictionless-market assumption he leans on (zero transaction costs, unlimited borrowing at one rate, free short selling) is closest to true for exchange members. For a retail account paying per-contract commissions on nine rehedges, the theoretical edge can be entirely consumed by the process of capturing it.
- It gives you no way to know what the right volatility is, and it says so. Every method in the book takes your volatility forecast as an input and tells you what to do with it. Nothing in the book tells you how to produce one you can trust. Chapter 20 is the closest attempt: weight historical windows by how well they match the option's time to expiration, notice that volatility mean reverts and is serially correlated, then use EWMA or GARCH. That is a description of the problem, not a solution to it. His preface is blunt: 'I make no claim to having found a magic secret to successful option trading. Anyone seeking such a formula will have to look elsewhere.' Take this book as an edge-generator and you will build beautifully constructed positions around a guess.
- The data ends around 2012, and the short-volatility lesson got very expensive after that. His conclusion in Chapter 20 is that implied volatility on the S&P 500 was normally too high, by up to 10 percentage points, and he explains it with an insurance analogy: buyers overpay for the rare catastrophe. That is the variance risk premium, and it is real. But readers who took it as a trade instruction met February 5, 2018, when the VIX more than doubled in a session and the XIV short-volatility note was terminated, and then March 2020. Natenberg does show the counterexample (on September 8, 2008, three-month implied volatility was 22 percent against a subsequent realized 72 percent) but he treats it as a year of extremes rather than deriving a drawdown or capital rule from it. His sizing advice, do it bigger when the margin for error is bigger, is a comparative rule with no absolute floor. The book will teach you that short volatility has positive expectancy. It will not teach you how much of it will end your account.
- There is nothing on same-day-expiry options, which are now most of the volume. The 2nd edition was written when the shortest realistic option had weeks to run. Chapter 8 rehedges weekly. Chapter 20 builds a term structure from two-month, four-month, six-month and eight-month expirations. Chapter 25's treatment of VIX futures ends with contango and backwardation charts from 2011 and 2012. Zero-days-to-expiration SPX options did not exist as a daily product and now dominate index option volume. Several of the book's core mechanics degrade inside a single session: scaling volatility by the square root of time assumes many periods, realized volatility measured settlement to settlement assumes a settlement, and mean reversion in implied volatility assumes there is time for it to happen. You can still use the Greeks. You cannot use the timeframe assumptions underneath the examples.
- Roughly a third of the book is not written for you, and the edition you buy may not be the one described here. The practical spine is Chapters 6 through 13, plus 20, 23 and 24. Chapters 15, 16, 19 and 22 (option arbitrage, early-exercise boundaries, the binomial derivation, index futures fair value) are professional market-making material and are dense. Natenberg also deleted the bibliography in this edition, so there are no onward pointers. On editions: everything cited here comes from the 2nd edition of 2015, which he wrote roughly twenty years after the previous one. The interest-rate examples still use 6 and 8 percent, so redo the arithmetic before applying anything rho-related or any early-exercise threshold. If you buy a used first edition to save money, you lose five chapters he added in 2015: forward pricing, the second-order Greeks (vanna, charm, volga), the Black-Scholes model itself, binomial pricing, and the entire volatility-contracts and VIX chapter.
- Every example is sized for an account you probably do not have. The positions in the book are 10, 35, 85, 100 and 200 contracts. The butterfly he recommends on risk grounds in Chapter 13 is a 100 x 200 x 100, a three-legged spread he himself flags as harder to execute and more costly in bid-ask terms. He describes a 100 x 50 ratio spread as possibly 'a small trade' to a well-capitalized trader. None of the risk-management logic in Chapter 13, which is the best chapter in the book, is reachable with a five-figure account: you cannot equalize theoretical edge across three candidate structures, you cannot leg into a butterfly at target prices, and you cannot rehedge nine times without commissions swamping the result. The reasoning transfers. The trades do not.
Who it is for
Buy it if
You already trade options and can compute a standard deviation without looking it up. You are the trader who keeps getting the direction right and losing money anyway, or who sells premium profitably for months and then gives it all back in one week. This book explains exactly why both of those happen, and Chapter 13 alone will change how you size.
Skip it if
You are looking for setups, entries, or a strategy to run. This book contains none and says so in the preface. Skip it if you have not yet traded options at all (start with the mechanics somewhere cheaper), if you trade a small account where multi-leg spreads and frequent rehedging are not economic, or if the math above the level of a spreadsheet average is going to stop you. Buying it because it is 'the standard text' and then reading three chapters is the most common way money gets wasted here.
As an Amazon Associate I earn from qualifying purchases. It costs you nothing extra, and it does not change what gets recommended or what the verdict says.