While we were not necessarily surprised by the Central Bank of Kenya (CBK) decision last week to keep its policy rate unchanged at 18 percent given the current and forward looking dynamics (short-end yields have narrowed sharply since January) to both inflation (Headline CPI 15.6 percent y/y in March from a cyclical peak of 19.7 last November, with food inflation—36 percent of the total basket—down to 22.1 from 26.2 and transport inflation moderating from 28 to 15.9 over that same period) and exchange rate appreciation (KES/USD up ~29 percent since mid-October partially on the back of timeframe restrictive currency restrictions relating to currency borrowing by offshore banks as well as final round policy rate hikes) we still find the overall tight monetary policy current to be subtly shifting underneath the surface as officials seem to be just as, if not more concerned with export (i.e. manufacturing) competitiveness than with underlying price pressures and currency stability (factoring in a weaker than expected Q4 2011, we expect overall economic growth rate of roughly 4 percent in 2011 versus 5.6 in the previous year as net exports act as a drag on consumer driven output). That said while our target rate per a Taylor’s rule analysis indicates the CBK could reasonably slash some 400bp from its benchmark rate through 2012 in order to buoy growth, we acknowledge the MPC continues to see potential upside risks to inflation given a deteriorating current account deficit (~12 percent of GDP in 2011 versus 7.8 in 2010) as well as sticky core figures—11.2 percent y/y compared with the government’s short-term target of 9 percent, a testament to rising crude prices (~20 percent of the country’s imports) given food’s previously stated disinflation—as well as above-target credit extension that taken together keep its policy mindset circumspect, especially given lingering La Niña effects as well as unusual tropical cyclone activity in the Indian Ocean that will likely delay long rains, impacting electricity generation and agricultural production and further underpinning food and fuel prices.[i]
Additionally, on a technical front USD/KES rests gingerly atop its 200dma support line and appears to have made a sort of rounding bottom or saucer pattern typically indicative of a reversal in price behavior. The pair now trades again roughly in line with the reaction highs of February, a level which also served as support one year ago while preceding an ultimately rapid ascent to the 106 level. Simply put, therefore, the pair is at a critical zone of truth whereby a definitive break one way or the other would likely be more than transitory, a view corroborated by historical volatility (HV), i.e. the standard deviation of day-to-day log price changes expressed as an annualized percent. For example, we examined a 5-day moving average of the ratio between the pair’s short term and long term HV, looking for spots whereby the relationship dipped below and then traded above .5, implying reversion to the mean volatility (we cite Professor Turan G. Bali’s 2006 study in particular regarding the mean-reverting behavior of stochastic volatility)[ii] and an impending price break. Interestingly the recent potential bottom in USD/KES coincided with an approximate 6-week stretch of low (<.5) relative volatility, and given that the average has again receded we would suspect a similar, albeit more sustained breakout looms. While given its recent history of protracted consolidation periods (see 2008-2009 and 2010-2011) we target a near term run to 84 in the coming month.
Trade recommendation: While mindful of said currency dynamics rising real rates (i.e. policy rate less headline inflation), which have now shifted into positive territory, should offset technical movements enough such that the sharp decline in the short-end (-157bp in March from the previous auction) is likely to give way to a more tempered rate of decline such that in the near term we see enough near-term downside, rather than upside risk to inflation as well as an adequately supportive currency environment (assuming authorities maintain current restrictions) to continue to recommend un-hedged local bonds.
[i] Bali, Turan G. and Demirtas, Ozgur K. Testing Mean Reversion in Stock Market Volatility. September 2006. Available at: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=936647