The 2013 Taper Tantrum: What Moved, and What Barely Did

Between May 2 and September 5, 2013, the 10-year US Treasury yield rose by roughly 137 basis points and the dollar gained nearly 9 percent against a basket of emerging-market currencies. The episode became known as the "taper tantrum," and it's often recalled loosely as a broad risk-off event. The data from the episode itself tells a narrower, more specific story, and the narrowness is the more useful lesson.

What actually happened, in three stages. The Federal Reserve's May 1 policy statement was the first step, introducing language about a potential future change in the pace of its bond-buying program, language that drew only a muted initial market reaction. The second and better-remembered step came three weeks later: on May 22, Federal Reserve Chair Ben Bernanke testified before Congress and, in response to a question, indicated the Fed could begin slowing its purchases in the coming meetings if the economy kept improving, the exchange most commonly cited as the taper tantrum's trigger. The third step came at the Fed's June meeting, where the post-meeting communication was read as confirming the tapering signal and coincided with a sharp jump in the 10-year yield. Measured across the full window from May 2 to September 5, the 10-year Treasury yield climbed about 137 basis points and the dollar appreciated by close to 8.8 percent against an equally weighted basket of 18 emerging-market currencies.

What barely moved at all. Over the same four-month window, the VIX rose by only about 2 points, a modest shift by the standard of a genuine risk-off episode, and high-yield credit spreads widened by only around 16 basis points, a small fraction of what shows up during episodes driven by actual concern about corporate defaults or systemic financial stress. Judged by those two measures specifically, this wasn't a broad flight from risk.

Why the split matters. Markets have more than one channel through which a shock can travel: a rates-and-policy-expectations channel, and a risk-sentiment channel, and they don't always move together. The evidence from 2013 points to the episode being dominated by the first channel rather than a broad US risk-premium shock, not that the second channel played no role at all. The Fed communication repriced the expected path of interest rates directly, and higher US yields changed the relative return available on US assets and tightened global financial conditions, part of why capital pulled back toward the US, while shifts in risk premia and country-specific vulnerabilities put additional pressure on several EM currencies on top of that. Little of that required investors to become more fearful about corporate defaults or systemic risk, which is consistent with why credit spreads and the VIX barely moved.

Why emerging markets bore a disproportionate share of the reaction. Countries running current account deficits and relying on foreign capital inflows to fund them were more exposed to a sudden pullback in that capital than countries with stronger external positions, a distinction that shows up clearly across the emerging-market currencies that sold off hardest during this period compared to those that held up better. The taper tantrum wasn't a uniform emerging-market event either, it exposed which countries had accumulated more external vulnerability during the preceding years of unusually easy global financial conditions. A current account deficit was one vulnerability among several, not a complete ranking model, inflation credibility, reserve adequacy, and existing foreign positioning all mattered too.

The lesson for reading a similar episode today. A market narrative that gets labeled with an emotionally loaded name like "tantrum" can retroactively get remembered as a broad risk-off panic even when the underlying data shows something narrower and more mechanical. Before assuming a shock is a systemic risk event, checking whether credit spreads and volatility measures actually moved, or whether the reaction is concentrated in rates and currencies instead, is a useful first step toward identifying which channel is actually doing the work.