The 2013 Taper Tantrum: How a Hint About Fewer Bond Purchases Repriced Global Markets

Key Takeaways

• The ten-year Treasury yield rose from 1.66 percent on 1 and 2 May 2013 to 2.98 percent on 5 September, a jump of about 132 basis points in four months, without the Federal Reserve raising a single short-term interest rate.
• The two-year Treasury yield barely moved by comparison, from 0.20 percent to a peak of only 0.52 percent over the same stretch, which is the cleanest evidence that this was a long-end, term-premium story rather than a story about the Fed tightening policy.
• The average 30-year fixed mortgage rate rose from 3.35 percent in the first week of May to 4.58 percent by the fourth week of August, according to weekly Freddie Mac data published through the Federal Reserve Bank of St. Louis.
• The Indian rupee fell from about 53.65 to a record 68.80 per dollar by 28 August 2013, a decline of roughly 28 percent; the Brazilian real fell about 22 percent over the same window, while the South African rand kept weakening past August and closed out the year down about 16 percent.
• On 18 September 2013 the Federal Open Market Committee surprised markets by leaving its $85 billion monthly purchase pace unchanged, stating outright that the tightening of financial conditions already under way, if sustained, could slow the economy's improvement.
• The Fed did not actually reduce its purchase pace until 18 December 2013, seven months after the testimony that triggered the sell-off, cutting to $75 billion a month effective January 2014; the unemployment rate had fallen from 7.5 percent in May to 6.7 percent by then.
• A survey of primary bond dealers taken immediately after the June 2013 meeting split almost evenly on timing: a third expected tapering to begin by September 2013, which did not happen, while more than a third of a separate panel of forecasters expected it to wait until 2014 or later.

What Happened During the 2013 Taper Tantrum?

By the spring of 2013 the Federal Reserve had been buying bonds on a large scale for more than four years, and the third round of that program, generally called QE3, had a feature the earlier two did not: no announced end date. When the Federal Open Market Committee launched it on 13 September 2012, the statement committed to purchasing agency mortgage-backed securities at $40 billion a month, and said the Committee "will closely monitor incoming information on economic and financial developments in coming months" rather than naming a stopping point or a total size. Three months later, on 12 December 2012, the Committee folded in a separate maturity-extension program that was expiring and added $45 billion a month of longer-term Treasury purchases, bringing the combined pace to $85 billion a month, a level that held for the next year and a half.

An open-ended program has an unusual property: the market cannot price its end from the announcement, because there is no scheduled end to price. Instead the market has to infer the end from every subsequent word Fed officials say about the economy. That inference problem broke open on 22 May 2013, when two things happened on the same day. First, the Fed released minutes of its 30 April to 1 May meeting containing a sentence that had not appeared before: a number of participants said they would be willing to reduce purchases "as early as the June meeting" if data showed sufficiently strong growth. Second, Bernanke testified to Congress's Joint Economic Committee, and in the question session, by contemporaneous press accounts, indicated the Committee could step down the pace of purchases within a few meetings if data cooperated. Coming from the same institution on the same day, the minutes and the testimony reinforced each other, and long-term yields, essentially flat through early 2013, began to climb.

The move accelerated on 19 June 2013, when the Federal Open Market Committee's post-meeting press conference gave the market something the May comments had not: an actual roadmap. Bernanke told reporters that if the economy behaved as the Committee expected, it would be "appropriate to moderate the monthly pace of purchases later this year," continuing "in measured steps through the first half of next year, ending purchases around midyear," at which point unemployment would likely be "in the vicinity of 7 percent." That is a specific, conditional, but concrete plan, and bond markets treated it as one. The ten-year yield jumped from 2.20 percent the day before the meeting to 2.60 percent within a week.

Chronology of the episode

Key dates with the ten-year Treasury constant maturity yield on the same trading day, from the Federal Reserve's H.15 statistical release.

Read as a whole, the shape is unusual for a page in this library. There was no failure, no bankruptcy, no bailout and no bank run. Every dollar the Fed had already committed to buying, it bought. The entire episode was a market repricing the value of a promise the Fed had not yet made, on the basis of hints about a promise it eventually did make seven months later.

What Exactly Did the Federal Reserve Say, and When?

Because this episode was caused entirely by language rather than by an action, the exact wording matters more here than in almost any other case in this library. Three documents carry the weight.

The minutes of 30 April-1 May 2013, released 22 May. Minutes are always released with a three-week lag, so this document described a meeting held three weeks earlier, but its release date is what moved markets, because it was the first time the public saw how divided and how close to action some officials already were. The relevant sentence: "A number of participants expressed willingness to adjust the flow of purchases downward as early as the June meeting if the economic information received by that time showed evidence of sufficiently strong and sustained growth." The same minutes recorded other participants preferring to wait for more evidence, so the document described a live disagreement, not a decision, but markets had not seen the word "June" attached to tapering before.

Bernanke's 22 May testimony to the Joint Economic Committee, same day. The Fed's own published testimony document is a prepared statement discussing the Committee's intent to keep assessing progress "in light of incoming information," without naming a timeline. The market-moving language came afterward, in the question-and-answer session, which the Fed does not publish as an official transcript. Financial media covering the hearing reported Bernanke indicating the Committee could step down its purchase pace within the next few meetings if data continued to improve, a characterization well corroborated by what happened to yields that same afternoon, but treated here as reported rather than as a verified quotation, since no official transcript of the exchange exists.

The 19 June 2013 press conference, the moment that actually set the terms. Here the language is unambiguous, because the Federal Reserve publishes its own transcript. Bernanke told reporters: "the Committee currently anticipates that it would be appropriate to moderate the monthly pace of purchases later this year. And if the subsequent data remain broadly aligned with our current expectations for the economy, we would continue to reduce the pace of purchases in measured steps through the first half of next year, ending purchases around midyear. In this scenario, when asset purchases ultimately come to an end, the unemployment rate would likely be in the vicinity of 7 percent, with solid economic growth supporting further job gains, a substantial improvement from the 8.1 percent unemployment rate that prevailed when the Committee announced this program."

The conditionality got lost. Every sentence in that roadmap is qualified: "if," "anticipates," "in this scenario." Bernanke was describing what the Committee would likely do under an economic path it expected, not a commitment independent of the data. Markets, and much of the subsequent commentary, treated it as closer to a fixed schedule. That gap between a conditional forecast and a perceived commitment is a large part of why the Fed spent the following three months trying to walk the tone back without reversing the substance, and why it ultimately reversed the September decision the market had priced in.

Notice what none of these three documents contains: a rate increase, a change in the target range for the federal funds rate, or an announcement that any purchase had actually been reduced. Every word quoted above describes a future intention conditional on future data. The entire sell-off that followed was the market pricing that intention before it happened, which is a mechanism worth naming precisely, since Federal Reserve policy rates and forward guidance explains why central-bank communication moves markets on its own, without any change in the policy rate itself.

How Far and How Fast Did Treasury Yields Actually Move?

The headline figure, a roughly 132 basis point rise in the ten-year yield between 2 May and 5 September, understates how unusual the move was, because it hides which part of the yield curve actually moved. A bond's price falls when the yield an investor demands rises, and the size of that price fall for a given yield move grows with the bond's duration, which is longer for longer-maturity bonds. That mechanism, covered in full in bond duration explained, is why a two-percentage-point move in the ten-year note wipes out years of interest income in a single quarter, while the same move in a two-year note barely registers.

Treasury constant maturity yields at three points on the curve, from the Federal Reserve's H.15 release.

Line the three columns up and the shape of the move becomes obvious. The two-year yield rose about 32 basis points from its low to its 2013 peak. The five-year rose about 120 basis points. The ten-year rose about 132 basis points. A move that grows with maturity, while the shortest maturity barely participates, is the signature of a shift in what bond investors call term premium, the extra yield demanded for the uncertainty of holding a longer bond, rather than a shift in the expected path of the short-term policy rate. The Federal Reserve's own December 2012 forward guidance, tying the funds rate to unemployment falling below 6.5 percent with inflation contained, kept the front end of the curve pinned near zero throughout, since nobody seriously expected 6.5 percent unemployment within two years in the spring of 2013. Investors did not think the Fed was about to raise rates. They thought the Fed was about to stop being the single largest buyer of longer-dated Treasury and mortgage securities, and repriced the securities it had been buying accordingly.

An independent Federal Reserve source corroborates the FRED data to the basis point: a 2017 speech by then-Vice Chairman Stanley Fischer, reviewing the episode, reproduced the daily ten-year yield series from the Fed's own H.15 release and marked the two catalytic dates directly on the chart, labeling 22 May 2013 "JEC testimony" and 19 June 2013 "June FOMC." That same chart shows the yield reaching 2.73 percent on 5 July 2013, matching the FRED series used throughout this page exactly, which is the kind of cross-check that belongs in any figure this page publishes.

Why Did Mortgage Rates React So Quickly, and What Slowed Down Because of It?

Fixed mortgage rates in the United States are priced primarily off longer-term Treasury yields and mortgage-backed security spreads, not off the federal funds rate, so a move concentrated in the ten-year note reaches a homebuyer's rate sheet within days rather than through the slower chain that a funds-rate change would take. Freddie Mac's weekly survey, published through the Federal Reserve Bank of St. Louis, shows the average 30-year fixed rate at 3.35 percent for the week of 2 May 2013, barely above the record lows of the prior year. By the week of 22 August it had reached 4.58 percent, a rise of 123 basis points in under four months.

Freddie Mac average 30-year fixed mortgage rate, weekly survey.

The single steepest week in Freddie Mac's full weekly series, a 53 basis point jump from 3.93 percent to 4.46 percent between 20 and 27 June, lands exactly on the week of the 19 June press conference, underlining how directly the mortgage market translates a Fed communication event into a household borrowing cost. A payment on a $300,000, 30-year loan at 3.35 percent runs about $1,322 a month in principal and interest; the identical loan at 4.58 percent runs about $1,534, a difference of roughly $212 a month that a household applying in August paid and one applying in May did not, for reasons that had nothing to do with that household's own creditworthiness. Refinance activity, running near record levels into early 2013 as homeowners locked in multi-decade lows, slowed sharply once the higher rates took hold. The FOMC's own September statement noted plainly that "mortgage rates have risen further," listing it as a headwind to the housing recovery the Committee was otherwise citing as evidence of progress, a rare case of a central bank naming, in an official statement, a cost its own words had helped create.

Why Did Emerging Markets Get Hit So Much Harder Than the S&P 500?

An investor tracking only United States stocks in the summer of 2013 would have had almost no reason to notice a crisis was underway anywhere. Robert Shiller's long-running monthly data on the S&P Composite, maintained at Yale University, show the index's monthly average price rising from about 1,640 in May 2013 to about 1,808 in December, a gain of roughly 10 percent across the very months the ten-year Treasury yield nearly doubled. United States equities were pricing an improving domestic economy, which is exactly what the Fed said justified slowing its purchases. Emerging-market currencies were pricing something else entirely: the withdrawal of a subsidy they had never been promised in the first place.

For the better part of four years, near-zero rates in the United States, Europe and Japan had pushed a large pool of global capital toward higher-yielding assets wherever it could find them, and emerging-market bonds, corporate debt and local-currency assets absorbed a large share of that flow. The trade underneath much of it was simple: borrow or fund in dollars at rates close to zero, buy assets denominated in currencies paying meaningfully more, and collect the spread as long as the currency did not move against you by more than the spread was worth. That profitability depends on two things staying stable: the size of the rate gap, and confidence that dollar funding stays cheap and abundant. Bernanke's 19 June remarks threatened both at once, making the trade less attractive and raising the cost of the dollar funding underneath every version of it, everywhere, simultaneously.

The result was not a crisis inside any single emerging economy's banking system, in the way this library documents for Asia in 1997 or the United States in 2008. It was a rapid, broad-based reversal of portfolio flows out of the economies whose currencies and government financing depended most heavily on those flows continuing, which is a narrower and more specific vulnerability than "emerging markets" as a category, and the next section works through why some countries were hit far harder than others.

Which Emerging Markets Were Hit Hardest, and Why Those Specifically?

Not every emerging-market currency moved by the same amount, and the pattern is instructive. Economies that had financed persistent current-account deficits, gaps between what they spent and earned abroad, by attracting exactly the kind of portfolio inflow described above, saw the largest currency declines, because their external financing needs did not pause just because the inflows financing them did. India, Brazil, Indonesia, Turkey and South Africa were singled out by market commentators through 2013 under a label that stuck, the "Fragile Five," because each combined a meaningful deficit with heavy reliance on portfolio capital rather than more stable, longer-term investment.

Currency depreciation against the United States dollar, measured from 1 May 2013 to each currency's own 2013 low, from Federal Reserve daily exchange-rate data.

Two things about that table are worth sitting with. First, the rupee and the real bottomed within days of each other, in the last two weeks of August, well after the initial 19 June shock, showing the pressure kept building for months rather than resolving in the first sharp reaction. The rand did not bottom with them: it kept sliding through the autumn and closed out the year at its weakest level, a reminder that "the taper tantrum" was not one synchronized event but a series of currency-specific reversals that happened to share a common trigger. Second, this page states three currencies rather than five, because Federal Reserve daily exchange-rate data used throughout this article does not cover the Indonesian rupiah or the Turkish lira on a comparable basis, and both are widely reported to have depreciated sharply over the same months. Rather than reach for a figure this page cannot verify against a primary source, that decline is described without a number: the rupiah and the lira both weakened substantially through the third quarter of 2013, and both countries' central banks responded with policy tightening before year end.

The mechanism connecting the deficit to the currency move is direct. A country running a current-account deficit spends more abroad than it earns abroad, and closes that gap with foreign capital, whether direct investment, bank lending, or portfolio flows into its bonds and stocks. Portfolio flows are the least sticky of the three: they can reverse in days. When the prospective return on holding rupee, real or rand assets narrowed relative to safer, more liquid Treasury securities, the marginal foreign holder had a reason to sell, and a currency financing a persistent external gap has no natural buffer against that reversal beyond its central bank's own reserves.

Why Did the Fed Surprise Markets by Not Tapering in September 2013?

By September 2013, most of Wall Street treated a reduction in the pace of purchases at the 17-18 September FOMC meeting as close to a foregone conclusion; June's roadmap had specifically pointed toward the second half of the year, and the economic data since then had not deteriorated. The Committee did not taper. Its statement gave the reasoning directly: "the Committee decided to await more evidence that progress will be sustained before adjusting the pace of its purchases," continuing purchases unchanged at $40 billion a month of mortgage-backed securities and $45 billion a month of Treasuries.

The more revealing sentence sat earlier in the same statement, describing the Committee's read of the economy: "the tightening of financial conditions observed in recent months, if sustained, could slow the pace of improvement in the economy and labor market." That is the Federal Reserve, in an official policy statement, naming the market's own reaction to its earlier communication as a reason to hold off on the action that communication had described. Mortgage rates were already more than a point higher than they had been in May. The Committee's judgment was that tapering on top of that self-inflicted tightening risked doing more damage to the recovery than the marginal $85 billion of ongoing purchases was doing good.

Two other pressures reinforced the caution, though neither appears as the stated reason in the statement itself. A partial federal government shutdown began at the start of October 2013 and a fresh debt-ceiling standoff ran alongside it, resolved only when Congress enacted the Continuing Appropriations Act, 2014, on 17 October, giving the Committee an added reason to avoid a second, self-generated source of uncertainty in the same window. Market expectations themselves were also genuinely split going into the meeting, a point the next section returns to with the actual survey data.

The September non-decision had an immediate market effect: the ten-year yield fell from 2.86 percent the day before the meeting to 2.69 percent the day of it, purely because a widely expected action did not occur. A market that reacts that sharply to the absence of an event is a market that had already priced the event as close to certain.

What Changed Between September and December That Let the Fed Finally Taper?

Three months later the Committee acted. Its 18 December 2013 statement announced that beginning in January 2014, mortgage-backed security purchases would fall to $35 billion a month from $40 billion, and Treasury purchases would fall to $40 billion a month from $45 billion, a combined reduction from $85 billion to $75 billion. Boston Fed President Eric Rosengren dissented, on the grounds that the reduction was premature given still-elevated unemployment and inflation running below target, a reminder that the decision was not unanimous even in December.

What had changed was mostly labor-market data and mostly the passage of time itself, which let markets absorb the June shock rather than react to it fresh. The unemployment rate, 7.5 percent in May 2013 when the episode began, had fallen to 6.7 percent by December, still above the Committee's 6.5 percent threshold but visibly closer to it, and closer to the "vicinity of 7 percent" Bernanke had described in June. The fiscal standoff that complicated September had been resolved by law in October, and the market had spent three additional months adjusting, so the December announcement, though the first actual reduction, produced a far smaller yield reaction than the May or June communications had: the ten-year yield closed at 2.89 percent that day, essentially unchanged from where it had spent most of the preceding two months.

The Committee simultaneously strengthened its guidance on when it would eventually raise the federal funds rate, stating it "likely will be appropriate to maintain the current target range for the federal funds rate well past the time that the unemployment rate declines below 6-1/2 percent," an explicit attempt to decouple tapering from any near-term rate increase. Purchases continued shrinking through 2014 and the program formally concluded at the Committee's 29 October 2014 meeting, the statement noting simply that "the Committee decided to conclude its asset purchase program this month." From the first hint in the May minutes to the final purchase, the full cycle ran about seventeen months.

Which Warning Signs Were Visible in Advance, and Which Only in Hindsight?

This episode offers an unusually clean test of how well professional forecasters could time a purely communication-driven event, because a survey exists that captured expectations at exactly the right moment: the week immediately following the 19 June press conference, when the roadmap was freshest.

Expected timing of the start of tapering, surveys taken 5-10 June 2013, reported in a 2017 Federal Reserve speech reviewing the episode.

Two findings sit inside that table. The primary dealers, who transact directly with the New York Fed's trading desk, put equal weight, 33 percent each, on September and December, essentially a coin flip between the wrong month and the right one. The broader Blue Chip panel put more weight, 36 percent, on 2014 or later than on any single 2013 quarter, and was wrong in the other direction. No single group's median forecast matched the actual date. That is not a criticism of forecasting skill; it is what an event driven entirely by a data-contingent policy path should look like, since the Fed itself had said the decision depended on data nobody, including the Fed, could see in advance.

Observable at the time. The size and open-ended structure of the purchase program itself, the numerical thresholds the Fed had already tied to future policy, the specific "around 7 percent" language from June, and the visible current-account deficits of the eventual Fragile Five were all public information before the sell-off's worst months. An investor did not need hindsight to know that an open-ended, unemployment-contingent program would eventually slow, or that certain emerging economies were funded by flows that could reverse.

Only clear in hindsight. The exact date remained genuinely unknowable in real time, as the survey data shows. So did the size of the reaction: nothing in the Fed's own communication predicted a 132 basis point move in the ten-year yield or a 28 percent currency decline from a hint about a still-distant, still-conditional policy shift. The magnitude was a market reaction to language, not a policy outcome the Fed itself could have forecast, which is precisely the distinction covered in cognitive biases in trading: knowing a mechanism exists is not the same as being able to price how far it will run.

Was This "Tightening"? Why Tapering and Raising Rates Are Not the Same Thing

Financial media coverage from 2013 onward has used "taper tantrum" and descriptions of Fed "tightening" almost interchangeably, and that habit blurs a distinction worth holding onto precisely because it recurs in every subsequent Fed transition.

Tapering means reducing the pace of new asset purchases. As long as that pace stays above zero, the Fed's balance sheet is still growing every month, which means monetary policy is still adding stimulus, just adding slightly less of it than the month before. The Federal Reserve's balance sheet did not shrink at any point during 2013; it grew every single month of the tantrum, and kept growing until the program formally concluded in October 2014. Tightening, in the sense investors generally mean the word, refers to raising the federal funds rate or actively reducing the balance sheet, both of which move short-term borrowing costs and bank reserves directly. Neither happened in 2013. The federal funds rate sat at zero to 0.25 percent on 2 May 2013 and sat at zero to 0.25 percent on 31 December 2013, unchanged for the entire episode.

What actually moved was market expectations about a future reduction in the flow of new stimulus, transmitted through the term-premium channel this page's yield table already documented: the front end of the curve, which prices the near-term policy rate, stayed anchored, while the long end, which prices the supply and demand for duration itself, moved sharply. That is a real and consequential market event, and this page does not minimize the size of the moves it produced. But calling it a tightening cycle, in the same sense as the deliberate rate increases this library documents in the 2022 rate shock or the sustained policy tightening of Volcker-era disinflation, mischaracterizes the mechanism, and mischaracterizing the mechanism is exactly what makes an investor expect the next Fed transition to look like this one when it may not.

Who Actually Lost Money, and Who Didn't?

No institution failed during the taper tantrum, no depositor lost access to funds, and no government had to intervene to prevent a default. The losses were real but they were mark-to-market losses on positions, not solvency events, and who absorbed them tracked closely with who held the longest-duration or least-liquid version of the trade that had been working for the previous four years.

Long-duration bond holders absorbed the most direct losses. Anyone holding intermediate or long-term Treasury or agency mortgage-backed securities in May 2013 saw the value of that holding fall through the summer, with the loss scaling with duration exactly as the yield table above shows. A pension fund or insurer running a long-duration bond ladder for liability matching absorbed a larger mark-to-market hit than a money-market investor holding paper that matures in weeks, even though both held instruments the market considered equally safe from default.

Emerging-market borrowers with dollar-denominated debt faced a second, compounding loss. A company or government that had issued debt in dollars, a common practice for issuers seeking lower rates than their local currency would command, found the local-currency cost of servicing that debt rising in step with the currency's decline, on top of any rise in the debt's own market yield. A borrower earning revenue in rupees or rand while owing dollars was exposed twice over, which is the same currency-mismatch dynamic this library documents at the sovereign level in the Asian financial crisis, though 2013's mismatches sat mostly with corporate and portfolio borrowers under floating exchange rates rather than with pegged currencies defended until they broke.

Leveraged carry trades lost the most, proportionally. An investor who had borrowed dollars cheaply to fund a levered position in higher-yielding emerging-market bonds experienced the currency loss, the bond-price loss, and a rising cost of the dollar funding itself, all at once, and often had to unwind at the worst possible moment because the same conditions that hurt the trade also tightened the terms on which it could be financed. That compounding is why currency and bond declines in economies with heavy foreign positioning tend to overshoot what the underlying data alone would justify.

United States equity holders and anyone holding short-duration cash were largely unaffected or better off. The equity gains documented above accrued to broad domestic stock holders throughout. And because the episode raised yields without any credit event, holders of very short-term instruments, whose yields began adjusting upward as the broader rate structure moved, had no downside from the repricing at all.

Why Is the Taper Tantrum a Poor Template for the Next Fed Transition?

Three features of 2013 were specific enough to the moment that expecting an identical replay the next time the Fed changes course is likely to be the wrong preparation.

Markets had never watched an open-ended program end before. QE3 was the first Federal Reserve purchase program without a pre-announced size or end date, so there was no historical template for how the Fed would communicate its way out of one, which is a large part of why every hint carried outsized information value. By the time the Fed built a second large post-crisis balance sheet during the pandemic and began reducing it in 2021-2022, both the Fed and the market had already lived through one full taper-and-runoff cycle, and communication around the second one, while still market-moving, produced nothing resembling a single-quarter, 132 basis point shock.

The starting level of yields was unusually low. A move from 1.66 percent to 2.98 percent more than doubled the ten-year yield's May starting point. The same absolute move starting from a higher base, as in more recent cycles, is a smaller proportional shock to bond prices and to the arithmetic of any yield-dependent trade built on top of it.

The emerging-market vulnerability was concentrated in a specific, nameable set of economies. The Fragile Five label existed because the vulnerability was genuinely narrow: large current-account deficits funded by portfolio flows. Several of those economies subsequently ran smaller deficits or built larger reserves, precisely because 2013 taught their policymakers what the vulnerability looked like from the inside. A future episode with a similar mechanism would very likely concentrate in a different set of countries than the 2013 list, identified by running the same balance-sheet questions again rather than by assuming the 2013 roster still applies.

Common Myths About the Taper Tantrum

"The Fed raised interest rates in 2013." It did not. The federal funds rate stayed at zero to 0.25 percent throughout the entire year. The event that moved markets was a shift in expectations about the pace of asset purchases, a different instrument entirely from the policy rate, as the earlier section on tapering versus tightening sets out.

"The Fed actually cut its bond buying in the summer of 2013." It did not. Purchases continued unchanged at $85 billion a month straight through the entire tantrum, including the September meeting where a reduction was widely expected and did not occur. The first actual cut came on 18 December 2013.

"Emerging markets crashed because their fundamentals were suddenly bad." The fundamentals that mattered, current-account deficits funded by portfolio inflows, were not new in May 2013; they had existed for years while capital kept flowing in on the strength of near-zero developed-market rates. What changed was the price of that capital's alternative, becoming more attractive as Treasury yields rose, a shift in relative pricing rather than a sudden deterioration inside any affected economy.

"Everyone should have known tapering was coming and priced it in advance." The survey data above shows professional forecasters genuinely split on timing even after the June roadmap made the eventual direction clear. Knowing a program would eventually taper is not the same as knowing when, and the size of the market reaction shows plainly that positioning had not already absorbed the news.

"The taper tantrum proves the Fed should never signal its intentions in advance." The Committee's own later behavior argues against that reading. Rather than retreating to less communication, the Fed spent the following decade refining forward guidance further, including numerical thresholds, dot plots and detailed press conferences, on the judgment that a market surprised by a lack of communication reacts at least as violently as one reacting to communication it partly misreads.

What a Reader Can Actually Carry Forward

The temptation is to reduce this case to "watch what the Fed says about tapering." That is too narrow, and it also assumes a reader is positioned to trade Fed communication professionally, which most are not. The more durable lessons sit one level up.

What generalizes

• A central bank's words are a policy instrument in their own right, separate from its actions. The entire 2013 move happened without a single rate change or a single reduced purchase for months. Anyone holding duration-sensitive assets, bonds, bond funds, or a mortgage they intend to refinance, is exposed to what a central bank says it might do, not only to what it has already done.
• Yield curve position matters more than "am I in bonds or not." The two-year note and the ten-year note are both "Treasuries," and one barely moved while the other moved by more than a full percentage point. Duration, not asset class, determined the size of the loss, which is exactly the distinction covered in short-duration versus long-duration bonds.
• External financing dependence is a specific, checkable vulnerability, not a vague "emerging market risk." A country, a company, or a fund's counterparty that relies on continued foreign portfolio inflows to roll over its financing is exposed to a change in the relative attractiveness of that capital's next-best alternative, which can shift for reasons that have nothing to do with the borrower's own conduct.
• Currency mismatch compounds interest-rate exposure rather than sitting alongside it. Dollar-denominated debt serviced with local-currency revenue turns a rate move and a currency move into one combined loss rather than two separate, smaller ones, which is a risk worth checking explicitly in any international holding rather than assuming diversification alone handles it.

What does not generalize

• The specific magnitude. A 132 basis point move in four months, starting from an unusually low yield level after an unprecedented open-ended purchase program, reflects a specific starting condition that will not recur identically.
• The Fragile Five roster. Which economies are most exposed to a portfolio-flow reversal changes as current-account positions, reserve levels and financing structures change; the 2013 list is a historical snapshot, not a standing watchlist.
• The seven-month gap between hint and action. That gap reflected the Fed's own deliberate caution after watching the market's reaction; a future central bank, having watched this episode, might communicate differently, faster or slower, in either direction.

The one question worth asking now

Not "when will the next taper tantrum happen," which nobody, including the institution doing the tapering, can answer with precision. The answerable question is narrower: for any bond holding, mortgage decision, or international position a reader holds, what happens to its value if the relevant central bank simply talks about doing less, without doing anything yet? That question requires no forecast of Fed policy, only an honest look at duration and financing structure.

References

Every figure on this page was verified against the following sources, each retrieved on 28 August 2026:

• Federal Reserve Board: Minutes of the Federal Open Market Committee, April 30-May 1, 2013: the quoted sentence on participants' willingness to adjust purchases downward as early as the June meeting.
• Federal Reserve Board: Chairman Ben S. Bernanke, Testimony Before the Joint Economic Committee, May 22, 2013: the date, committee and prepared-remarks content of the testimony; this page does not quote the question-and-answer session, which the Fed does not publish as an official transcript.
• Federal Reserve Board: Transcript of Chairman Bernanke's Press Conference, June 19, 2013: the exact quoted language on moderating the pace of purchases, ending purchases around midyear 2014, and the 7 percent and 8.1 percent unemployment figures.
• Federal Reserve Board: FOMC Statement, September 13, 2012: the launch of QE3 at $40 billion per month of agency mortgage-backed securities, open-ended.
• Federal Reserve Board: FOMC Statement, December 12, 2012: the addition of $45 billion per month of Treasury purchases and the 6.5 percent unemployment / 2 percent-plus-a-half inflation forward-guidance thresholds.
• Federal Reserve Board: FOMC Statement, September 18, 2013: the decision to await more evidence before adjusting the pace of purchases and the explicit reference to tightening financial conditions.
• Federal Reserve Board: FOMC Statement, December 18, 2013: the reduction to $75 billion per month, the revised forward guidance language, and Eric Rosengren's dissent.
• Federal Reserve Board: FOMC Statement, October 29, 2014: the conclusion of the asset purchase program.
• Federal Reserve Board: Accessible Data for Vice Chairman Stanley Fischer's "Monetary Policy Expectations and Surprises," April 17, 2017: the independently sourced ten-year yield chart cross-checking the FRED series, the 22 May and 19 June date labels, and the Primary Dealer versus Blue Chip Economic Indicators survey data on expected taper timing.
• U.S. Government Publishing Office: Public Law 113-46, Continuing Appropriations Act, 2014: the October 17, 2013 enactment date resolving the government shutdown and debt-ceiling standoff.
• Robert J. Shiller, Yale University: U.S. Stock Markets 1871-Present and CAPE Ratio: the monthly average S&P Composite values for May and December 2013 used to describe the direction of United States equities during the episode.
• Federal Reserve Bank of St. Louis, for every yield, mortgage rate, exchange rate and labor-market figure quoted: 10-Year Treasury Constant Maturity Rate, Series DGS10, 5-Year Treasury Constant Maturity Rate, Series DGS5, 2-Year Treasury Constant Maturity Rate, Series DGS2, 30-Year Fixed Rate Mortgage Average, Series MORTGAGE30US, India / U.S. Foreign Exchange Rate, Series DEXINUS, Brazil / U.S. Foreign Exchange Rate, Series DEXBZUS, South Africa / U.S. Foreign Exchange Rate, Series DEXSFUS, and Civilian Unemployment Rate, Series UNRATE.

Figures deliberately not stated. This page gives no daily or month-end closing level for the S&P 500 in 2013, because the FRED series most pages on this site use for that index only retains a rolling ten-year window and does not reach back to 2013; the Shiller monthly-average series cited above is used instead and is explicitly labeled as a monthly average, not a daily close. This page gives no verified quotation from Bernanke's 22 May 2013 question-and-answer session, because the Federal Reserve's own published testimony document contains only the prepared remarks. It states no specific dollar figure for the Brazilian central bank's 2013 currency-intervention program, no specific interest-rate decisions by the Reserve Bank of India, Bank Indonesia or the Central Bank of Turkey, and no current-account-deficit percentage for any of the Fragile Five economies, because none of those figures could be confirmed against a primary or institutional source in this session. Where the mechanism is well documented but a specific magnitude is not, as with the Indonesian rupiah's and Turkish lira's 2013 declines, the direction is described and no number is supplied.

Method note: figures described as computed, including basis-point changes and percentage currency declines, were derived by Swoopr Investment from the daily series named above, retrieved on 28 August 2026. Treasury and exchange-rate figures are daily closing quotations from the Federal Reserve's H.15 and H.10 statistical releases as republished by FRED; the mortgage rate is Freddie Mac's weekly survey average, not a daily figure, so it is presented by survey week rather than by calendar date. The unemployment rate is the seasonally adjusted monthly civilian rate. Percentage currency declines are calculated from the 1 May 2013 level to each currency's own low within 2013, expressed as the increase in the dollar price of that currency, and are not annualized or otherwise adjusted.

This is educational content about a historical episode. It is not investment advice, it is not a forecast, and nothing here should be read as a claim about how the Federal Reserve or any currency will behave in the future.

Related Reading

• Market History Case Studies: where a pure communication shock like this one sits against episodes involving an actual institutional failure.
• The 2022 Rate Shock: the closest comparison in this library for an actual, deliberate tightening cycle, and the sharpest contrast with 2013's expectations-only move.
• The Asian Financial Crisis: currency mismatch and capital-flow reversal on a much more severe scale, under pegged exchange rates that eventually broke rather than the floating currencies that merely depreciated in 2013.
• Silicon Valley Bank and the 2023 Regional Banking Stress: the same duration mathematics applied to a single bank's balance sheet rather than to a sovereign yield curve.
• Volcker-Era Disinflation: what a deliberate, sustained rate-tightening campaign looks like, for contrast against 2013's single communication shock.
• Bond Duration Explained: the mechanism behind why the ten-year note moved so much further than the two-year note.
• Short-Duration versus Long-Duration Bonds: how to think about that same duration exposure when building a bond position today.
• Federal Reserve Policy Rates and Forward Guidance: how central-bank communication moves markets independent of any actual policy change.
• Yield Curve, Term Premium and Recession Signals: a fuller treatment of the term-premium concept this page's yield-curve section relies on.
• International Investing: how currency exposure interacts with, rather than simply adds to, an international bond or equity position.

Frequently Asked Questions