Recap 2014-09-29


Interesting bit from JPM on how asset classes react to forecast changes:

  • We do find evidence that it is possible to use past forecast changes to make profitable investment decisions. This is likely due to a lagged market reaction to our forecast changes, and to a modest amount of serial correlation in forecast changes. We note that selling 10-yr bonds when our economists have raised their growth forecasts over the past 1-4 weeks, and buying them when they lowered growth was a great strategy over the past 12 years with an average return to risk of 1.0.
  • Buying stocks in countries where growth has been upgraded and selling them where it has been downgraded lost money though. It is our belief that this is likely because higher growth led to higher bond yields and currencies, both of which are negative for stocks. Selling stocks where growth is being downgraded, a rather uncomfortable strategy, actually made decent money.
  • Across countries, we found it quite profitable to buy the currencies of the countries where we just upgraded growth (rising forecast revision index, FRI) against those we downgraded, or upgraded by less. We note that selling bonds of the rising FRI countries against the falling or less rising FRI countries only made money within the DM. Overweighting the stock markets with the higher FRI changes against the lower ones also made good money, but over half the return came from the currency. On a currency hedged basis, this cross country equity strategy performed OK, but was not stellar.
  • Generally, markets react relatively quickly to our forecast changes, typically within the month, but in cross country strategies, we find also that forecast changes over several months have impact, likely as it takes time to gauge diverging growth trends across countries.


  • Pending Home Sales declined -1.0% in Aug vs -0.5% exp and +3.3% prev
  • US Core PCE was stable at 1.5% vs 1.4% exp
  • German HICP was stable at 0.8% vs 0.7% exp
  • RBNZ’s intervention data for August reveals net NZD selling of NZ$521mn. This is the most significant open selling position we have seen from the RBNZ in several years. New Zealand prime minister John Key (a former FX trader) would like to see the kiwi at 0.65.
  • Pro-democracy protesters vowed to press ahead with demonstrations unless Hong Kong’s top official steps down, with thousands of people surrounding government offices after violent clashes paralyzed the city center. Protesters dressed in black have gathered in the Admiralty district to demand free elections, while blocking a main road into the central business area. Thousands of pro-democracy protesters are blocking Hong Kong’s streets, shutting down the territory’s business hub and ignoring appeals to leave. Crowds remained on the streets overnight after a day that saw riot police deploy tear gas and batons in a bid to disperse them.
  • Ibovespa futures plunged after a poll showed increased support for President Dilma Rousseff’s re- election bid ahead of the Oct. 5 vote. In the latest Datafolha poll released Sept. 26, Rousseff had 40 percent support in the first round, 13 percentage points ahead of Silva. Neves’s support was 18 percent. Datafolha surveyed 11,474 people from Sept. 25-26, and the poll has a margin of error of plus or minus 2 percentage points. Silva and Rousseff had 30 percent and 37 percent respectively in the previous Datafolha poll, published Sept. 18. h/t Jordan


  • Mon: NZ Building Permits, UK GfK Consumer Confidence, Japan Jobless Rate, NZ Business Confidence
  • Tue: Month End, UK House Price, EU Employment, EU CPI, Canada GDP, ChicagoPMI, US Consumer Confidence, Japan Mfg PMI, AU Retail Sales
  • Wed: EU PMI, US ADP, ISM, Australia New Home Sales, Building Approvals, Trade Balance
  • Thu: ECB, US Jobless Claims, Australia Service PMI, China Non-Mfg PMI, Japan Service PMI

2 thoughts on “Recap 2014-09-29

  1. Sir,

    How do you interpret the 5y5y breakeven curve? My understanding is that this represents where the market sees 5 year inflation in 5 years from now, but I am struggling to understand what it means when this diverges from the 10y spot breakevens.

    Thanks for your time

  2. Hi Ben – the 5y5y forward breakeven inflation rate is typically interpreted as the long run inflation expectation rate for the economy. 5 years has historically covered a cycle, and 5 years forward usually strips out the effects of expectations limited to the current cycle. The calculation is a function of the 10y and 5y rates, so a divergence between the 5y5y and the 10y will be explained by the spot 5y rate. Hope that helps!

Comments are closed.