Revisiting The “Tokyo Whale”

One of the things I’ve noticed in my (often spectacularly unsuccessful) attempts to explain to average investors the extent to which the Bank of Japan intervenes in domestic equity and bond markets, is that there’s a tendency for people to believe that what I’m telling them is too absurd to be true.

And while it’s often frustrating when people refuse to believe something that is not only explicit policy, but is in fact discussed all of the time and reiterated at televised policy meetings, I can almost forgive the retail crowd for their willful ignorance. Ignorance, after all, is bliss, and there’s nothing particularly blissful about coming to terms with the fact that the BoJ owns 75% of the Japanese ETF market.

Or maybe there is something blissful about that realization. After all, it should provide investors with some measure of comfort to know that a whale of that size is not only keeping a constant bid under Japanese equities, but is in fact timing purchases to coincide with dips.

Still, no one who fancies themselves an investor wants to admit they were so naive that they didn’t realize the sheer scope and outright brazen character of Kuroda’s intervention. It’s kind of like how no Trump supporter wants to admit that they didn’t realize their vote was going to end up plunging America into an Erdogan-esque experiment in autocracy.

Well, I’ve got to head out to dinner (appropriately enough at a local sushi joint), but before I go, I wanted to highlight a few charts and some brief commentary from SocGen regarding the “Tokyo Whale.” While none of this will be news to anyone who mans a sellside desk, I imagine it might be quite surprising for some of HR‘s retail audience (who knows, maybe some buyside folks don’t even appreciate how absurd this has become).

Without further ado, here’s SocGen…

The “Tokyo Whale” in the ETF market…

The term “whale” is frequently used to qualify big money participants in the market. The Bank of Japan has often been qualified as the “Tokyo Whale” since it became a major participant and holder in the Japanese government bond (JGB) market. Recently, this situation was replicated in the ETF market. At end-March 2017, BoJ’s cumulative ETF purchases were ¥13.1tn, with Japanese equity ETFs subject to BoJ purchases totalling ¥21.3tn assets. Based on the price performance of the Nikkei 225 and Topix indices and as dividends are not reinvested in Japanese equity ETFs but distributed, we can estimate the mark-to-market value of historical purchases. We assume the BoJ’s current ETF holdings stand around ¥15.7tn ($144bn), i.e. approximately 75% of the total assets in Japanese equity ETFs (orange line on below chart).

Eligible ETFs are physically replicated, which implies the ETF providers hold the underlying index constituents. From the estimated BoJ ETF holdings and index constituents list, we can deduce how much of the Japanese equity market is indirectly held by the BoJ through these ETFs. 

… but also, indirectly, in the Japanese stock market

BoJ implicit holdings are quite small at the index level but significant on some specific stocks

Comparing the BoJ’s ETF holding amounts per benchmark to each index’s market capitalisation is not relevant as many companies are common to the three (or two) indices. We have totalled the estimated amounts held for each stock through the three benchmark exposures and concluded that the BoJ may indirectly hold around 3.2% of the Nikkei 225 market capitalisation, 2.0% of the TOPIX and 3.1% of the Nikkei 400. These figures may seem quite low compared to the respective 75% and 40% estimated BoJ ownership of the Japanese ETF and government bonds markets. The impact at the stock level, however, can vary greatly from one stock to another.

Estimating the impact of BoJ ETF holdings on a stock’s free float

The calculations of a stock’s free float can vary significantly across market data providers, index providers and stock exchanges. This is the case with Japanese stocks in particular. However, these different sources agree on one point: the shares held by mutual funds including ETFs and Japanese trust banks are considered as free-float. As a result, the stocks held by the ETFs purchased by the BoJ are treated as free-float, which may be open to debate as these shares are not available for trading and the BoJ may hold its ETF shares over a long period. From this standpoint, we consider the shares owned by the ETFs in proportion to the ETF holding rate by the BoJ as stagnant and exclude them from our free-float estimates. To calculate these “ex-BoJ” free-float figures, we subtract the estimated BoJ indirect stock holdings from the free-float provided by Bloomberg — which appears to be less restrictive than that of the Tokyo Stock Exchange (TSE). The 15 companies for which the BoJ’s implicit holdings are the largest as a percentage of the stock market capitalisations are listed below. At end-March 2017, based on our calculations, the highest market cap the BoJ was indirectly holding was Advantest Corp at 14.3% and it held above 10% of seven companies. This means Advantest saw its “effective” free float market cap decline from 77.1% to 62.8% due to the BoJ policy, all things being equal. The company was not among the most heavily weighted stocks in the Nikkei 225 or TOPIX but had a relatively small market capitalisation compared to its weight in the Nikkei 225. Because the Nikkei 225 is a price-weighted index, the stocks that are the most exposed to ETF purchases (i.e. the most heavily weighted) are not necessarily the ones with the largest market cap and this is where the liquidity issues may come from.