The last two steps of the BOJ ETF exit plan proposed by Tomoyuki Shimoda are (3) a repository institution to be set up by the government would redeem the ETFs for the underlying cash equities and repackage them into index ETFs and (4) the repackaged ETFs would be sold to institutional investors (but not retail investors) at the index’s average intraday price on the sale date (see Part 2 for a discussion of the first two steps).

The key point here is that the repackaged ETFs would be sold solely to institutional investors. The ETFs owned by the BOJ trade on the Tokyo Stock Exchange (TSE). As such, they can be purchased even by retail investors. Shimoda presumably proposed step (3) to create institutional ETFs distinct from existing ETFs. He recommended that the repackaged ETFs be sold at their underlying index’s intraday average price (step (4)) to incentivize institutional investors to buy as much of them as possible. On market up days, indexes’ closing prices should usually be above their respective intraday average prices. If so, institutional investors would benefit by buying the repackaged ETFs at a discount to their closing price. Conversely, there typically should be no repackaged ETF sales on market down days because no one would want to buy at a premium to the closing price. Shimoda’s proposed scheme would allow the market’s appetite for the ETFs to drive their sales volume instead of imposing minimum quarterly sales quotas.

The next question is, What is institutional investors’ propensity to actually buy ETFs? Based on the TSE’s annual ETF Beneficiary Survey (ownership survey), life insurers and pension trusts (≒ pension funds) respectively owned ¥770bn and ¥300mn of ETFs that track Japanese equity indices as of July 31, 2021. By comparison, life insurers and pension funds’ Japanese public equity holdings at December 31, 2021, totaled ¥23trn and ¥16trn, respectively, according to the BOJ’s flow-of-funds statistics.

Institutional investors’ ETF holdings are minuscule relative to their Japanese public equity holdings. This huge disparity presumably reflects Japanese institutional investors’ preference for discretionary investment contracts over ETFs as a passive investment vehicle. Discretionary investment contracts offer advantages that ETFs cannot match, including lower management fees and more control over voting of shares. Kenichi Hirayama also noted these advantages in his book I discussed in Part 1 of this series.

The statistics cited above are limited to institutional investors in Japan but even when foreign institutional demand is taken into account, institutional demand for the repackaged ETFs would likely be considerably lower than the BOJ’s total ETF holdings of ¥35trn.

The Shimoda proposal does not require all ¥35trn to be sold. If institutional demand is lacking, the repository institution would continue to hold the ETFs. Shimoda likely realizes that such an outcome may be inevitable. If so, his proposal, which appears to be a divestment plan at first blush, would actually be more of a freeze.

Shimoda’s proposal is certainly “market friendly,” as he put it. Additionally, the BOJ would be shedding a lot of market risk exposure by offloading its ETF holdings to a repository institution (the government). While such an outcome would seemingly be a win for all concerned and pretty much resolve the BOJ’s ETF exit conundrum, our discussion does not end here. When other perspectives are factored into the equation, the Shimoda proposal may not be the optimal way forward. (To be continued)

• Katsutoshi Takehana

  Group Manager

  Financial Market & Digital Business Research Department