Skip to content
Menu
Finabase Learn

Articles · Great investors

George Soros and reflexivity: the Quantum Fund and Black Wednesday

How George Soros ran the Quantum Fund, what his theory of reflexivity says, how the 1992 bet against sterling worked, and the losses that belong in the same record.

On this page
  1. Who Soros is
  2. Reflexivity in plain terms
  3. Black Wednesday: the trade step by step
  4. The record, including the losses
  5. What investors can take from it
  6. How Finabase helps
  7. Frequently asked questions
  8. Summary
  9. Check your understanding

George Soros (born 1930) is the best-known speculator of the late twentieth century, and the reason is one day: 16 September 1992, when the United Kingdom gave up defending the pound inside the European Exchange Rate Mechanism. Soros's Quantum Fund had bet heavily that it would, and the profit was reported at about $1 billion. Newspapers named him "the man who broke the Bank of England," a title that overstates his role but stuck.

The trade was the visible part of a longer story. Soros ran a fund that, by the figures it reported, compounded at a rate few investors have matched for three decades. He also had a theory of markets, reflexivity, which he thought explained both booms and busts, and he took losses that were as large as his wins were famous. This article explains the theory in plain terms, walks through the sterling trade step by step, and sets out the record, the bad years included. It is a history, not a template: almost nothing in it can be repeated by an individual investor.

Who Soros is

Soros was born in Budapest in 1930, survived the Nazi occupation of Hungary as a teenager, and moved to London in 1947, where he studied at the London School of Economics. The philosopher Karl Popper, who taught there, shaped his thinking more than any economist did: Popper argued that our knowledge is always fallible, and Soros spent much of his life applying that idea to markets.

He moved to New York in 1956 and worked as an analyst and arbitrage trader. In 1969 he started a fund called Double Eagle for the firm Arnhold and S. Bleichroeder, with about $4 million of capital. In 1973 he and his analyst Jim Rogers left to run it independently as Soros Fund Management; the main fund was renamed Quantum in 1979. Rogers left in 1980. In 1988 Stanley Druckenmiller joined and ran Quantum day to day for the next twelve years.

Reflexivity in plain terms

Standard economics treats prices as reflecting fundamentals: earnings, interest rates, supply and demand. Investors may misjudge them for a while, but the errors are random and prices are pulled back to value. Soros disagreed with the second part. In The Alchemy of Finance (1987) and later papers he argued for two linked ideas.

  • Fallibility. Participants never have a complete or correct picture. Their views are always partly wrong.
  • Reflexivity. Those views do not just observe the market, they act on it. When people buy because they expect prices to rise, prices rise, and the higher prices can change the underlying reality.

Soros called the first channel, reality shaping what people think, the cognitive function, and the second, what people think reshaping reality, the participating function.

Reflexivity: views move prices, and prices move the fundamentals Participants’ views Expectations, always partly wrong (Soros: “fallibility”) Act on them: buy, lend, sell Prices Share prices, credit terms, exchange rates Set by what people do Fundamentals Earnings, collateral values, borrowing capacity, policy Change because prices move participating function prices feed back cognitive function Illustration of the feedback loop Soros described in The Alchemy of Finance (1987). In a boom, rising prices improve the fundamentals (more collateral, cheaper capital), which seems to confirm the bullish view, until the gap between belief and reality can no longer be sustained.
Illustration of the reflexive loop described in The Alchemy of Finance (1987).

Lending booms are the clearest case. Banks lend against property; the lending pushes up property prices; higher prices make the collateral look more valuable, which justifies more lending. The fundamentals (the value of the collateral) are not independent of the belief (that property is a safe thing to lend against). Each round seems to confirm the last until the gap between belief and reality becomes too wide, and then the same loop runs backward: falling prices destroy collateral, banks pull back and prices fall further. The The 2008 financial crisis: housing, Lehman, TARP and the 57% fall followed that pattern closely, and Soros used it as his main example in a 2008 book.

Reflexivity is closer to a way of looking than to a model with testable equations. Economists have pointed out that much of it overlaps with existing ideas, such as feedback trading, credit cycles and herding. Soros himself said the theory did not tell him when to act. Its practical value for him was a habit: look for situations where a widely held belief is propping up the very conditions that make it look true, and ask what happens when it stops.

Black Wednesday: the trade step by step

The setup

The Exchange Rate Mechanism (ERM) was a system that held European currencies within bands around agreed rates against each other, a forerunner of the euro. The United Kingdom joined on 8 October 1990 at a central rate of 2.95 Deutsche marks per pound, with the pound allowed to move up to 6% either side. That put a floor of roughly DM 2.77–2.78 under sterling.

Defending a fixed or banded rate means keeping enough demand for the currency. The tools are buying it with reserves and raising interest rates so that holding it pays more. In 1992 both were costly. German interest rates were high because the Bundesbank was fighting inflation after reunification, so UK rates had to stay high too, while Britain was in recession. Markets doubted that the government would keep paying that price.

The asymmetry

This is the part that made the trade attractive, and it is the most useful idea in the story. A speculator who sold sterling short against marks near the floor faced two outcomes:

  • If the pound held, it could not fall much below the floor, and it was unlikely to rise far above it either. The loss on a short position was limited to a few percent plus the interest cost of being short a high-yielding currency for a while.
  • If the pound left the band, nothing held it up. It could fall 10% or more.

Druckenmiller has said in interviews that the idea was his and that Soros's contribution was to insist the position be much bigger. Accounts put Quantum's short sterling position at about $10 billion, built up over August and September 1992 (contemporary press reports; the fund never published its trades).

The day

On the evening of 15 September, remarks by the Bundesbank president Helmut Schlesinger, reported in the press, suggested that a broader realignment of European currencies was needed. Traders read it as a signal that Germany would not help sterling. Selling began early on the 16th.

Sterling in the ERM, October 1990 to 17 September 1992 Oct 8, 1990 UK joins the ERM at DM 2.95, ±6% band Sep 15, 1992 Bundesbank chief’s remarks reported Sep 16, am Base rate raised 10% → 12% Sep 16, pm 15% announced, never applied Sep 16, eve UK leaves the ERM; rate held at 12% Sep 17 Base rate back to 10% Published dates and rates (Bank of England, HM Treasury). The pound then fell well below its old ERM floor.
Published dates and rate decisions, October 1990 to 17 September 1992.

The Bank of England bought pounds with its reserves all morning. The government raised its base rate from 10% to 12%, then announced a further rise to 15% in the afternoon. Neither stopped the selling; the 15% rate never took effect. That evening the chancellor, Norman Lamont, announced that Britain was leaving the ERM, with the rate held at 12%; it was back at 10% the next day. Sterling fell well below its old floor over the following weeks.

Quantum's profit was reported at around $1 billion (some accounts say £1 billion), and the fund made more on related positions in other European currencies and interest rates. HM Treasury estimated in 1997 that the episode cost the UK about £3.1 billion, revised to about £3.3 billion in papers released in 2005, most of it in the form of profits the government would otherwise have made on its reserves.

What it did and did not show

Soros did not cause the devaluation. Many banks, companies and funds sold sterling that day, and the economic strain on the rate was real. What the episode showed was a reflexive trap: the more the market doubted the peg, the higher rates had to go to defend it, and the higher rates went in a recession, the more the market doubted it. The policy meant to prove commitment became evidence that it could not last. Many economists later argued that leaving the ERM helped the UK, which cut rates, adopted inflation targeting and recovered in the following years.

The record, including the losses

Hedge funds are private, and their published records come from the funds themselves or from investors and journalists who saw the letters. Nothing below is audited by Finabase; treat each figure as reported.

Period Result as reported Reported by
End 1969–end 1980 Fund up 3,365% versus 47% for the S&P 500 Company history (fund figures)
October 1987 crash Lost about $200 million in a day and at least $350 million Company history; press reports vary
1992 At least $1 billion from the sterling short Press; Soros interviews
1998 Large losses in Russia; Quantum still ended the year up Company history
Early 2000 Down about 22% for the year by late April, on tech CNNMoney, 28 April 2000

The 1987 loss is the clearest example of a reflexive thinker misjudging a reflexive market. Soros had expected a crash in Japan rather than the US, was long US stock index futures going into October, and sold near the bottom of Black Monday 1987: the 22.6% crash and why circuit breakers exist. The 2000 loss came after Quantum bought technology stocks late in the The dot-com bubble: the 1995–2002 boom and the 78% Nasdaq fall; Druckenmiller left that April. In 2011 Soros Fund Management returned outside investors' money and became a family office, managing the Soros family's and his foundations' assets. Soros has since given much of his fortune to the Open Society Foundations; this article covers only his investing.

A widely quoted figure puts Quantum's average return above 30% a year over its three decades as a public fund. It comes from the fund's own reporting and cannot be independently checked, so treat it as an indication, not a measured record.

What investors can take from it

Very little of Soros's method transfers directly. He traded currencies, bonds and index futures with heavy leverage, had direct access to policymakers and other large traders, and could afford to be wrong in billion-dollar sizes. Three ideas do transfer:

  • Look for asymmetry. The sterling trade was large because the downside was small and the upside large, not because Soros was certain. The same question, "what do I lose if I am wrong?", applies to any position. See Position sizing: how many shares to buy for a set risk.
  • Expect feedback, not just mean reversion. Prices can push fundamentals further from "normal" for a long time. Bubbles and crashes are not just mistakes that markets correct quickly. Mean reversion vs trend following: two rules on one market covers the trading side of the same question.
  • Admit error fast. Soros was known for reversing positions abruptly when the facts changed. His bad years were the ones where he did not, or did so at the worst moment.

How Finabase helps

You cannot replay a 1992 currency trade in Finabase, but you can study the markets the story involves. The indices page shows the FTSE 100 and other world indexes live, and a long-run chart such as SPY at maximum range shows the 1987 and 2000 episodes on real data. The scanner lists the day's largest moves, which is where reflexive surges and collapses tend to show up first. For currency mechanics, read The forex market explained: how currency trading works and Currency risk in a portfolio: how exchange rates change returns.

Frequently asked questions

How much did Soros make on Black Wednesday?

About $1 billion, as widely reported at the time and later confirmed in interviews, with more from related positions. The fund did not publish its trades, so the exact figure is not public.

What is reflexivity in simple terms?

It is the idea that what investors believe changes prices, and prices then change the reality the beliefs are about, for example by making credit cheaper or collateral more valuable. The loop can push markets far from equilibrium in both directions.

Did George Soros break the Bank of England?

Not alone. He held one of the largest short positions, but many institutions sold sterling, and the underlying cause was the gap between UK economic conditions and the interest rates needed to hold the ERM rate. The phrase is a newspaper label, not an economic finding.

Did the Quantum Fund ever lose money?

Yes. It lost heavily in the October 1987 crash, took losses in Russia in 1998 and was down about a fifth by late April 2000. Its long-run record as reported is still exceptional.

Summary

George Soros built the Quantum Fund into one of the most profitable hedge funds of its era, by its own reported numbers, and became famous for the 1992 bet that the United Kingdom could not keep sterling in the ERM. His theory of reflexivity holds that beliefs move prices and prices reshape fundamentals, which is why booms and busts can run far beyond what value alone would justify. The sterling trade worked because the downside was capped and the upside was not, which allowed a very large position. The same record contains a costly misjudgment in 1987 and a late bet on technology in 2000, a reminder that even the best-known speculators are often wrong and that the size of a bet matters as much as the idea behind it.

Check your understanding

4 questions on this lesson. Each answer comes with the reason, and nothing is scored.

  1. Why could Quantum make its sterling short so large in 1992?

    1. Soros knew in advance that the UK would leave the ERM
    2. The ERM band limited how far the pound could rise against the mark, so the loss if sterling held was small, while a break could produce a large gain
    3. Short positions in currencies cannot lose money
    Show the answer

    B. The ERM band limited how far the pound could rise against the mark, so the loss if sterling held was small, while a break could produce a large gain The trade was asymmetric. Near the band floor, being wrong cost a few percent plus carry; being right was worth 10% or more. A small, capped downside is what justified the size.

  2. Which situation is the clearest example of reflexivity?

    1. A company beats its earnings estimate and the stock rises
    2. Rising house prices make collateral look more valuable, banks lend more against it, and the extra lending pushes prices up further
    3. A stock falls after a bad quarter and later recovers to its average valuation
    Show the answer

    B. Rising house prices make collateral look more valuable, banks lend more against it, and the extra lending pushes prices up further In reflexivity the belief changes the fundamentals it is about. Lending against rising collateral feeds the price rise that made the collateral look safe, until the loop reverses.

  3. On 16 September 1992 the UK raised its base rate to 12% and announced 15%. Why did that fail to stop the selling?

    1. Higher rates in a recession made the policy look unsustainable, so markets doubted the government would keep paying the price
    2. The rate rises were too small to matter to any investor
    3. Interest rates have no effect on exchange rates
    Show the answer

    A. Higher rates in a recession made the policy look unsustainable, so markets doubted the government would keep paying the price Defending the peg with ever higher rates during a recession signalled strain, not strength. The defence itself became evidence that the rate could not last.

  4. What should you make of a reported figure such as "Quantum averaged over 30% a year for three decades"?

    1. It is an audited, independently verified record
    2. It is a self-reported hedge fund figure: an indication of an exceptional record that outsiders cannot check, and it says nothing about the large losing years inside it
    3. It proves that macro trading beats index funds for most investors
    Show the answer

    B. It is a self-reported hedge fund figure: an indication of an exceptional record that outsiders cannot check, and it says nothing about the large losing years inside it Hedge fund records come from the funds or their investors. The same record includes the 1987 and 2000 losses, and one manager's result says nothing about what most investors would earn.

Sources and further reading

  1. Soros — The Alchemy of Finance, 1987 (reflexivity; the "real-time experiment" of 1985–86)
  2. Soros — Fallibility, reflexivity, and the human uncertainty principle, Journal of Economic Methodology 20(4), 2013
  3. Bank of England — Official Bank Rate history (changes of 16 and 17 September 1992)
  4. HM Treasury — Papers on Black Wednesday released under the Freedom of Information Act, 2005 (cost estimate)
  5. Soros Fund Management LLC company history, International Directory of Company Histories (Quantum returns 1969–1980; 1987 and 1992 figures)
  6. CNNMoney — Soros fund managers step down, 28 April 2000
  7. Wikipedia — Black Wednesday (timeline and Treasury cost estimates, with references)