The Bank of Japan has for years bought Japanese stocks as part of its massive monetary easing program to lift the country out of deflation and hit a 2% price-stability target.

Under governor Haruhiko Kuroda's quantitative and qualitative monetary easing (QQE) plan, stock-buying through exchange-traded funds started in 2013 at a pace of about 1 trillion yen annually, expanding to about 3 trillion yen in October 2014, and further to about 6 trillion yen in July 2016.

The purpose was to increase aggregate demand and thus inflation, as well as to encourage Japanese savers to take on more risk by buying equities.

But QQE has not yet delivered the hoped-for results, with inflation (excluding food and energy) stubbornly stuck around 0% despite the BOJ's huge efforts.

Meanwhile, stock purchases have helped raise equity prices but they have made the central bank a very large investor, raising problems of stock valuation and corporate governance. It is time to think again -- and consider phasing out the stock purchases.

There are very few central banks in the world that have purchased domestic stocks on this scale and for such a long time (for more than five years under QQE but for nearly eight years since its original introduction in late 2010) as part of the conduct of monetary policy.

The current ETF purchases are also different from when the BOJ previously purchased bank stocks to protect financial stability -- 2 trillion yen in 2002-2004 (in a domestic banking crisis) and 400 billion yen in 2009-2010 (during and after the global financial crisis).

So, when the BOJ embarked on mass stock buying, it had no history to draw on. Similarly, the extremely challenging task of normalizing policy and ultimately selling stocks purchased is without precedent.

Stock prices began to rise from late-2012 in anticipation of the BOJ's aggressive monetary easing (see chart). They increased for various reasons, including monetary easing, yen depreciation and strong corporate profits, partly boosted by the BOJ's policy. Externally, higher U.S. stock prices and the appreciation of the U.S. dollar against major currencies also contributed.

But problems emerged. First, the hoped-for increase in Japanese individuals buying shares failed to materialize as they remained highly risk-averse. Instead, Japan's stock market has been increasingly dominated by foreign investors as a group. (see chart)

Second, the BOJ's large-scale purchases have turned the central bank into a top investor, second only to the Government Pension Investment Fund and Blackrock. Since the BOJ does not exercise voting rights, the bank's position as a silent investor may adversely affect corporate governance reforms now being pursued by the Japanese government.

Some listed firms may enjoy higher stock prices than they otherwise would and attract more investors while feeling less pressure to improve management and business strategies.

In the worst case, this may delay necessary corporate restructuring, undermine labor productivity growth and hurt potential economic growth.

Finally, ETF buying may have not only reduced downside risk on all stocks purchased but also lead to the possible overvaluation of some small-capitalization stocks.

This is because the central bank's buying of ETFs focused on tracking the main indices -- the Nikkei Stock Average, or Nikkei 225, the Tokyo Stock Price Index, and the JPX-Nikkei 400. Some small-cap stocks included in the Nikkei 225 may have received a particularly large extra price boost.

While the BOJ adjusted its purchases to take account of such risks, the problems remain. Small-cap shares will likely face a sharp sell-off when the BOJ begins to reduce ETF purchases.

Given that the likelihood of achieving 2% inflation is low, because underlying inflation is weak (consumer price inflation was at just 0.2% in August 2018 after excluding all food and energy), the BOJ may find it necessary to revise its QQE. In the process it could unwind ETF purchases.

With the possible phasing out of the program in mind, the central bank in July 2018 decided to adjust ETF purchases -- and buy more flexibly depending on stock market conditions, while sticking to the official annual purchase plan of about 6 trillion yen.

This suggests the BOJ intends to cope with the side-effects of its buying and possibly prepare for "stealth tapering" -- the quiet phasing out of the purchases -- as it is with its government bond-buying program.

Indeed, the BOJ reduced ETF purchases immediately starting in August to about 238.5 billion yen, compared with the previous month, which translates into 2.9 trillion yen on an annualized basis.

But let's not exaggerate -compared with the same month in the previous year, the increase was still 5.8 trillion yen only modestly below the 6 trillion yen average.

Tapering -- that is, reducing -- the amount of ETF purchase must be carefully conducted. The BOJ currently holds ETFs of about 21 trillion yen based on the acquisition price The potential impact on stock prices remains large.

For the time being is the central bank should tacitly begin to reduce the annual pace of ETF purchases toward about 3 trillion yen -- the level before the 2016 expansion, while maintaining the official annual target of 6 trillion yen. It is better to reduce ETFs now when corporate profits are high and the economy is expanding than in adverse conditions.

A cut in ETF purchases toward zero may be challenging since the ETF purchases are part of the QQE program.

Before taking any clear steps toward monetary policy normalization, the BOJ should introduce flexibility in interpreting the 2% price stability target. It should incorporate the 1% upper and lower range (plus or minus 1%) of the target into the 2% target itself. The BOJ and the government would then not need to abandon the 2% target while in practice aiming at a 1% figure that would be acceptable to the public.

Such flexibility would be better than lowering the target from 2% to 1% since many central banks have begun to interpret the 2% price stability more flexibly. For example, Sweden's Riksbank adopted plus or minus 1% target range into the 2% target in 2017. Moreover, a scheduled increase in the consumption tax in October 2019 from 8% to 10% is expected to generate inflation above 2% for a year. That gives officials the opportunity to introduce a 1% to 3% inflation target range that could be easily presented to the public.

Once the price stability target becomes more flexible, the BOJ may be able to reduce ETF purchases more clearly. It could also then expand the target range for yields on 10-year bonds further from the current plus or minus 0.2% range and reduce Japanese government bond purchases to the level of net issuance (about 20 trillion yen annually).

Completing the process of tapering out purchases of ETFs and bonds, and eliminating the 10-year yield target may take much longer since the Japanese economy may face an economic slowdown after the 2020 Tokyo Olympic Games.

So, the full normalization of monetary policy, notably raising short-term interest rates, is difficult to forecast. It will be some time before the BOJ can follow the U.S. Federal Reserve. But it could and should start reducing the distortions caused by stock buying much earlier.

Sayuri Shirai is a professor at Keio University, a visiting scholar to ADB Institute, and a former Bank of Japan board member.