A previous post argued that the MPC would need to set any unemployment rate threshold below 6.5% for markets to believe that current official rates will be held until 2015-16. It didn’t and they don’t.
The two-year-ahead forward rate on government liabilities – a gauge of market expectations of overnight rates in two years’ time – closed at 104 basis points yesterday, up from 90 bp last Tuesday, before the MPC’s forward guidance announcement. It is rapidly approaching a June peak of 121 bp – see first chart.
The spread between the two- and one-year forward rates is now 74 bp, above a June high of 71 bp. This suggests that markets expect official rates to be hiked by three-quarters of a percentage point between August 2014 and August 2015.
Labour market statistics released yesterday are consistent with the suggestion here that the unemployment rate will breach the 7.0% threshold by mid-2014, reflecting stronger GDP growth and slower productivity expansion than forecast by the MPC.
The rate was stable at 7.8% in the three months centred on May but the single month reading fell to 7.4% in June, a level also recorded in March, with April and May both at 8.0% – second chart. Put differently, the four- and two-month moving averages fell to 7.7%.
Unemployment must decline by about 250,000 to reach 7.0%. The claimant count – available two months before the official measure – fell by an average of 25,000 per month in the three months to July. This pace of decrease, if sustained, suggests that 7.0% will be hit in 10 months.
The fall in the claimant count is consistent with vacancies and survey data signalling a strong pick-up in labour demand. Vacancies surged by 8% in the six months to July, an increase historically consistent with six-month growth in employee numbers of 1% or more. 1% expansion equates to 250,000 new workers.
Strong demand for new workers casts doubt on the view of the MPC’s productivity optimists that firms have hoarded labour so will not need to recruit as the economy picks up.
The claimant measure captures only about 60% of total unemployment but may be a superior guide to trends because it is a comprehensive count. The official measure, by contrast, is based on a survey and is subject to significant sampling error. The Office for National Statistics estimates sampling variability at +0.3 percentage points (95% confidence interval).
The third chart shows two models designed to “predict” the official unemployment rate based on the claimant count; the second model includes a time variable to allow for trend divergence between the two measures. The models are used to project paths for the official measure assuming that the claimant count continues to fall by 25,000 per month.
The July claimant count reading is consistent with an official unemployment rate of 7.4% according to the first model; the second suggests 7.9%. Based on the assumed decline, the first model – the more optimistic – implies that 7.0% will be hit before end-2013. The second model, which incorporates a recent tendency for the claimant count to fall relative to the official measure, predicts August 2014.
