The U.S. rates research team of Bank of America/Merrill Lynch reasons that fears of less accommodative monetary policy can trigger a rise in U.S. Treasury yields that goes beyond the rationally expected path in fed funds rates. Catalysts of such non-fundamental dynamics can be (i) increased mortgage convexity risk, (ii) spillovers and repercussions from other bond markets, and (iii) duration hedging in the wake of bond fund outflows. Thereby, large institutional flows in a market with few players to warehouse risk can lead to an overshooting of yields.
Higher UST yields: Is this fundamental or technical?, Bank of America/Merrill Lynch, 11 June 2013.
(Original research is proprietary and non-public, please contact Bank of America/Merrill Lynch for access)
“Carry trades have dominated the fixed income landscape for the last few years as low [U.S.] Treasury rates incented investors to take on more duration, convexity or spread risk. Low rates and a range bound market also reduced the incentive to hedge for higher rates (since it would reduce returns). The [May/June 2013] move in rates to highs not seen since 1Q 2012 and the rise in volatility have reduced the incentive to hold on to carry trades (worse Sharpe ratio) and increased the need to hedge. Unfortunately, the ability of dealers to warehouse the shedding of risk is much more limited due to the regulatory environment, which argues for an exaggerated price move to these flows.”
BAML research focuses on three key transmission factors in that process:
Mortgage convexity risk
“[Market practitioners define mortgage convexity risk as the impact of active risk management of mortgage portfolios on the overall fixed income markets. Due to the sheer size of the mortgage market, the fixed income markets are net short volatility to homeowners in the form of the prepayment option.] Convexity hedging flows tend to exacerbate rate sell-offs especially in the 5-10 year part of the curve, widen swap spreads and increase short dated volatility.”
“The crucial factors when considering mortgage hedging risk is the move in Treasury rates and the sensitivity of the primary mortgage rate to the rise in Treasury rates. In the 20bp increase in Treasury rates that occurred from May 1 to May 9, primary mortgage rates rose only by about 7bp. In the 30bp rise in 10y Treasury that followed, primary mortgage rates rose by almost as much (28bp). This sensitivity of primary rates to Treasury rates incorporates the spread between current coupon MBS rate and Treasuries (mortgage spread) and the spread between the current coupon MBS rate and primary mortgage rate (primary-secondary spread).”
“MBS spreads…[have] been driven by fears of tapering of QE3, which impact demand dynamics for MBS [in the same direction as swap and treasury yields]. The primary-secondary spread…at about 85bps…does not allow originators to absorb any rise in mortgage rates…Therefore, in our view, any increase in Treasury rates could trigger significant convexity hedging needs… We estimate paying needs of USD145bn [1% of U.S. GDP] 10-year equivalents in a 25bp rise in primary mortgage rates.”
The repercussions of higher Japanese Government bond (JGB) yields
“There has always been a strong correlation across global bond markets between JGB, bunds and Treasury yields…Increased volatility in the JGB market increases the risk of a VaR shock among Japanese banks similar to 2003…Between mid June and end July 2003, global bond markets suffered their worst losses since 1994 despite mixed economic data- 10y USTs by 130bps, 10y JGBs by 50bps and 10y Bunds by 70bps.
Bond fund outflows
“May  was one of the worst months in terms of performance across fixed income indices. This performance could be a catalyst for further outflows, which would be a risk for credit as well as duration. US High yield funds and intermediate government funds had the largest outflow on record [in early June]. Mortgage funds also have had significant outflows over the last two weeks…These funds could be looking towards the Treasury market (given much better liquidity) to shed duration.”