More like zip than ZIRP

Yesterday's monetary policy easing by the Bank of Japan was presented in some press reports and blog posts as a return to the zero interest rate policy (ZIRP) that Japan tried from 1999 to 2000, and again from 2001 to 2006. No doubt the press were guided by the official post-decision statement, which described the policy as a "virtually zero interest rate policy". RebelEconomist is not convinced.

To be precise, the new policy is to guide the inter-bank uncollateralised overnight call rate to a range of "0 to 0.1%", compared with its previous target of "around 0.1%" maintained since December 2008. However, as footnote 2 of yesterday's statement states, the interest paid on excess reserves under the BoJ's complementary deposit facility (originally introduced in October 2008) remains at 0.1%. As described in detail in an earlier post here, paying interest on reserves effectively sets a floor to interest rates in the inter-bank market, because it pays any bank with a reserve account at the central bank to accept loan offers below the interest rate paid on reserves and deposit the money with the central bank. Depending on the significance of inter-bank market participants without access to remunerated reserves, transactions costs and operational constraints (such as timing differences between the close of the inter-bank market and the latest transactions allowed with the central bank), arbitrage should not permit inter-bank interest rates to fall much below the rate paid on reserves. And indeed, a US Federal Reserve study of foreign central banks' experience with remunerated reserves by Bowman, Gagnon and Leahy published in March 2010 found that since the complementary deposit facility was introduced in Japan, the overnight call rate had never traded below 8 basis points. This suggests that yesterday's BoJ policy change represents a negligible easing in terms of short-term interest rates.

There was also another easing measure announced by the BoJ yesterday. The BoJ undertook to purchase outright ¥5tn (about $60bn) of longer-term government and shorter-term private sector bonds including commercial paper (CP), asset-backed commercial paper (ABCP), corporate bonds, Japanese Real Estate Investment Trusts (J-REITS) and exchange-traded funds (ETFs), partly to reduce term and credit/liquidity spreads respectively, and partly of course to supply additional base money in pursuit of the lower overnight call rate target. Of this ¥5tn, "about ¥3.5tn" ($42bn) is to be allocated to government bonds and treasury bills, and only "about ¥1tn" ($12bn) to CP, ABCP and corporate bonds. Compared with the BoJ's existing holdings of JGBs (presently ¥80tn or $960bn), the outstanding stock of JGBs (¥734tn or $8.3tn as of end-June 2010), or the total size of the Japanese bond market (about $12tn or ¥1100tn according to the BIS as of end-March 2010, of which $9.8tn comprised JGBs) the planned purchases represent modest amounts. A rough estimate of the likely effect on yields of these additional acquisitions can be derived by comparison with the estimated effect of the Fed's treasury purchases under its own quantitative easing program, as the size of the US treasury market ($10.0tn at end-March 2010) is of a similar size to the (then) $9.8tn JGB market. According to a recent Federal Reserve discussion paper by D'Amico and King, the Fed's purchase of $300bn of treasuries in 2009 generated a sustained fall in mid-curve treasury yields of about 50bps. Assuming a similar impact of BoJ purchases on the JGB market, the $42bn JGB purchase announced yesterday could be expected to reduce mid-curve yields by about 7bps only.

Overall therefore, yesterday's BoJ easing move is modest, and may have been designed as much as a concession to the BoJ's critics in government and industry as an attempt to stimulate the Japanese economy.