800 Years to Price Risk

01一本讲「风险」如何被发明的书

In 1996, Peter Bernstein published Against the Gods: The Remarkable Story of Risk.

The subtitle contains a word — remarkable. It's about something we take for granted today but is actually extraordinary: how humans learned to put a price on risk.

We now think it's natural that risk can be calculated — insurance, options, credit, portfolios all rest on the premise that risk is quantifiable. But for most of human history, that premise simply didn't exist.

The ancient Greeks had the best mathematics, but they never computed probabilities. Why? Because they believed the future was determined by the gods — calculating probability was an affront to them. The idea that risk can be calculated is itself a rebellion against the notion that fate is determined by the gods. That's what the title Against the Gods means.

02最关键的转折:概率论的诞生

The book's core is the story of how probability theory was born from gambling.

In 1654, a French gambler asked mathematician Blaise Pascal a problem about splitting a pot in an unfinished game. Pascal corresponded with Fermat, and they accidentally laid the foundation of probability theory. The origin of modern risk management is two mathematicians solving a gambler's problem.

There's something delicious about that — the most serious tool in finance came from the least serious activity: gambling.

Then came Bernoulli with expected utility — the same amount of money has different value to a poor man vs. a rich man, so decisions shouldn't be based on amounts alone but on utility. This is the direct origin of modern risk aversion.

Then Gauss invented the normal distribution (the bell curve), allowing measurement errors and volatility to be quantified. Then Bayes — a method for updating old beliefs with new information.

This entire chain — from gambler to Pascal to Bernoulli to Gauss to Bayes — is the complete story of how humanity learned to manage uncertainty. Every financial tool we use today stands on some link in this chain.

03为什么这段历史对投资者重要

The biggest takeaway from this book isn't knowledge — it's perspective. It shows you that the tools you use every day are young, fragile, and likely flawed.

VaR (Value at Risk) models use the normal distribution — but financial markets are not normally distributed; their tails are much fatter than the bell curve suggests. Those "once-in-a-century" events of 2008 should, according to the normal distribution, happen once every tens of thousands of years, but they occurred several times in a decade. The model isn't wrong — the assumption that markets follow a normal distribution is wrong.

That's the warning Bernstein gives at the end of the book — The danger of risk management tools lies not in their imprecision, but in the illusion they create that you have risk under control. LTCM had two Nobel laureates and the most sophisticated models, yet it blew up in 1998. Why? Because their models assumed a milder randomness than the real world.

The more precise the risk model, the bigger the safety illusion it can produce — this is the deepest lesson this book leaves for investors.

04我跟伯恩斯坦不同的地方

First, his optimism about quantification was partly falsified by 2008.

The book was written in 1996, the golden age of financial engineering — derivatives, quant models, risk management tools were booming. Bernstein's tone is optimistic: humanity is getting better at taming risk.

But 2008 proved that these tools failed precisely when they were most needed. Bernstein in 1996 couldn't fully foresee that the proliferation of quant tools would itself create new systemic risk — everyone using the same models, all selling simultaneously in a crisis, amplifying the collapse. Taleb later articulated this clearly; Bernstein didn't.

Second, he underestimated how much behavior undermines probability.

Bernstein's narrative arc is "the progress of rationality" — humans getting better at calculating probabilities. But behavioral finance (Kahneman, Thaler) shows that even when people can compute probabilities, they don't act on them. Loss aversion, overconfidence, anchoring — these create a chasm between "can compute" and "will use." Bernstein didn't give this chasm enough attention.

Third, his perspective is too "Western rationalist."

The book frames the calculability of risk as a triumph of human reason, an advance of civilization. But this narrative has a blind spot — it assumes the world is essentially computable, we just haven't computed enough yet. Taleb, Mandelbrot, and others would disagree: some risks are fundamentally incomputable, no amount of rational progress can capture them. Bernstein stands with the "optimistic computability" camp — it's a stance, not a settled truth.

Fourth, it offers little operational guidance for ordinary investors.

This book is an intellectual history, superb at that, but it almost never tells you "how to manage your own risk tomorrow." It helps you understand where risk concepts came from, but it doesn't give you tools. That's not a flaw — it's a matter of positioning. But readers should know: this is a book for understanding, not for operating.

05伯恩斯坦 vs 塔勒布:可计算 vs 不可计算

This book and Taleb's work frame the most fundamental debate about risk.

Bernstein's position: Risk can largely be computed and managed. Humanity is progressing along this path, and that's a civilizational achievement.

Taleb's position: The risks that truly matter (tail events, black swans) are fundamentally incomputable. Any confidence that "I've got it calculated" is dangerous.

Who's right? My take: Bernstein is right in normal times; Taleb is right in extreme times.

99% of the time, markets follow roughly computable patterns, and risk management tools work. But your fate is determined by that 1% of extreme moments — when all computable models fail, and only a Talebian structure of "you can afford to lose" can save you.

My own posture: Use Bernstein's tools for everyday risk management; adopt Taleb's attitude for tail risk protection. You need both because they address two different kinds of risk.

06写在最后

Peter Bernstein was more than a writer — he was a Wall Street veteran who managed money, taught, and ran one of the most influential investment journals. He was 77 when he wrote this book.

What moved me most about this book isn't its intellectual density — it's its humility. Bernstein spent a lifetime writing about risk, but at the end of the book, he doesn't boast that "we've conquered risk." He says: We invented tools to manage risk, but we have never, and will never, truly conquer uncertainty.

The title Against the Gods contains a kind of tragic grandeur — humanity has been fighting the unknowability of fate, winning many local battles (insurance, options, portfolio theory), but the war is never winnable.

That posture is the healthiest one for an investor: neither abandoning tools to manage risk (which is resignation) nor believing tools can eliminate risk (which is arrogance).

Walking the tightrope between those two is what every mature investor does for a lifetime.

Bernstein spent 800 years of history to show you how that tightrope came to be. That's the greatest value of this book.