Japan Investment Report 2019 — The Gap Between Assets and Price

Japan presented an unusual proposition through this period: a large, developed, politically stable market where a substantial share of listed companies traded below the accounting value of their net assets.

That condition requires explanation. In an efficient market, a company worth less than the sum of its assets should attract a buyer who acquires it, sells the assets and captures the difference. The persistence of the condition across many companies and many years indicates something preventing that mechanism from operating.

The cause was capital allocation. Many Japanese companies held very large cash balances and other non-operating assets relative to their size. Cash earns essentially nothing, particularly in a zero-rate environment. A company with substantial idle cash has a mechanically depressed return on equity, because the denominator includes capital that produces no return.

The market's response was rational: value the operating business reasonably and apply a discount to the idle assets, on the assessment that they would never be deployed productively or returned to shareholders. A yen of cash that will never be distributed is not worth a yen to a minority shareholder.

Why the condition persisted is the more interesting question, and the answer is structural:

• Cross-shareholdings. Companies held shares in each other — customers, suppliers, banks — creating a bloc of shareholders who supported management for relationship reasons rather than financial ones.
• Boards had limited independence, with membership drawn substantially from within the company.
• The market for corporate control was inactive. Hostile acquisition was rare and culturally difficult, so the mechanism that normally disciplines poor capital allocation did not operate.
• Cash was viewed as prudence, and after decades of deflation and financial stress, a large balance was regarded as sound management rather than as an inefficiency.

The catalyst was institutional rather than market-driven. Governance codes were introduced and strengthened, requiring independent directors and explanations of capital policy. Exchange pressure on companies trading below book value followed. Cross-shareholding unwinding accelerated, encouraged by disclosure requirements.

Why idle cash depresses value more than it should

The arithmetic is worth setting out, because it explains both the discount and why it is larger than a naive view suggests.

Return on equity is profit divided by shareholders' equity. Equity includes all assets net of liabilities — operating assets and idle cash alike.

Cash earning nothing raises the denominator without raising the numerator, so ROE falls. A company with a good operating business and a large cash pile reports a mediocre ROE, and is valued as a mediocre business.

A worked illustration:

A company with ¥100bn of operating assets earning ¥15bn of profit has a 15% return on those

assets — a good business. Add ¥100bn of idle cash and equity becomes ¥200bn while profit is

unchanged. Reported ROE is 7.5%. The operating business has not changed. The reported return

has halved.

If the company returned the cash, equity falls to ¥100bn and ROE returns to 15% — with no

operational change whatsoever.

*Derived: illustrative arithmetic with round numbers, ignoring interest income and tax.

Not a description of any actual company.*

The market's discount goes further, and this is the part that produces sub-book valuations. If investors believe the cash will never be returned, they apply a discount to it. A company holding ¥100bn that will never be distributed is worth less than a company holding ¥100bn that will be — because the minority shareholder can only realise value through distribution or through a sale of their shares to someone who expects one.

So the discount compounds: the ROE depression makes the business look mediocre, and the assessment that the cash is permanently trapped discounts the cash itself. Together they can push a company below book value.

The gap is not a mispricing of the assets. It is an accurate pricing of the probability that a minority shareholder will ever see them.

This is what makes the governance reform thesis coherent. The gap closes not through the assets becoming more valuable but through the probability of distribution rising. A company that credibly commits to returning capital has changed the second variable without touching the first.

Why the discipline mechanism was absent

The persistence of the condition for decades requires explaining, and the explanation is about ownership structure.

In most developed markets, three mechanisms discipline capital allocation:

• The market for corporate control. A company allocating capital poorly becomes an acquisition target, because a buyer can capture the value the incumbent is not.
• Activist investors. A shareholder can build a stake and campaign for change, supported by other shareholders whose interest is financial.
• Board oversight. Independent directors representing shareholders challenge management on capital policy.

All three were weak in Japan, for related reasons:

• Cross-shareholdings created a stable, supportive shareholder base. A company whose shares are substantially held by its customers, suppliers and lenders has shareholders whose primary interest is the commercial relationship, not the share price. They vote with management.
• Hostile acquisition was rare, both culturally and practically — with a supportive shareholder bloc, an acquirer cannot assemble a majority.
• Boards drawn from within the company had limited independence from the management they oversaw.

The cross-shareholding is the load-bearing element. Remove it and the other two mechanisms begin to operate: a company without a supportive bloc is acquirable, and a shareholder base composed of financial investors will support a campaign for better capital allocation.

That is why cross-shareholding unwinding is the indicator to watch, and it is disclosed. It is a better measure of whether the reform is real than any statement of intent, because it changes who votes.

The catalyst was institutional

The change, when it came, arose from institutions rather than from markets — which is unusual and worth noting.

Governance codes established expectations for board composition, independent directors, and disclosure of capital policy. These operate on a comply-or-explain basis, which means companies must either meet the standard or publicly justify not doing so. Public justification is itself a pressure, because it makes the deviation visible and comparable.

Stewardship expectations for institutional investors required them to engage with companies and to disclose voting. This matters because it changed the behaviour of domestic institutions, which had historically supported management by default. An institution required to disclose its votes and explain its engagement cannot passively support poor capital allocation.

Exchange pressure on companies trading below book value — requiring them to publish plans for improving capital efficiency — created a specific, measurable and public obligation.

Why an institutional catalyst works differently from a market one:

• It applies broadly rather than company by company, so it changes the whole market's behaviour rather than picking off individual targets.
• It is slow, operating over years rather than through a discrete transaction.
• It is more durable if it persists, because it changes expectations rather than a single company's ownership.
• It can be reversed if the pressure weakens, which is the principal risk.

For an investor the implication is about timescale. A thesis dependent on institutional change is a multi-year thesis, and the relevant question is not whether the change is happening but whether the pressure will persist long enough for it to compound.

A return source that does not require growth

The most useful feature of the Japanese proposition is what it is uncorrelated with.

Most investment cases require growth. A company grows revenue, grows earnings, and the share price follows. That growth depends on economic conditions, on demand, on competitive position — all of which correlate across markets.

The capital efficiency case does not require growth. A company that returns excess cash, unwinds cross-shareholdings and improves its return on equity creates shareholder value without growing at all. The return comes from the gap between current and potential capital efficiency, not from expansion.

That makes it genuinely uncorrelated with the drivers dominating the rest of the region — trade conditions, technology supply chains, the dollar cycle, Chinese demand. A portfolio holding Japanese capital efficiency exposure alongside regional growth exposure holds two things driven by different variables, which is what diversification is supposed to mean and rarely does.

The specific risks are equally identifiable:

• Superficial compliance. Plans published and not implemented. The measurable test is capital actually returned, not plans announced.
• Pressure weakening. The catalyst is institutional and could be relaxed.
• Currency. As the 2022 regional report describes, a foreign investor's return can diverge substantially from the market's. The Japanese equity return and the foreign investor's return are separated by the yen, and the separation has been large.
• Timescale. Governance change is slow, and a thesis requiring years is difficult to hold through periods when it is not working.

Why the discount persisted for decades

A gap this large and this measurable, in a developed market with sophisticated participants, should not persist for decades. That it did is the most interesting fact about the situation, and the explanation is worth setting out because it identifies what would have to change.

The standard mechanism by which such gaps close requires someone with both the incentive and the ability to act. An acquirer buys the company, deploys or returns the idle capital, and captures the difference. That mechanism requires the acquisition to be possible.

Why it was not possible:

• The shareholder register was not for sale. Cross-shareholdings meant a substantial portion of shares were held by parties whose primary interest was a commercial relationship. They would not sell to a hostile acquirer, and they would vote with management. An acquirer cannot assemble a majority from a register that is largely committed.
• The social cost of a hostile approach was high, which deterred domestic acquirers and complicated foreign ones. This is not a legal barrier but it operated as one.
• The activist route was equally blocked. A campaign for better capital allocation requires other shareholders to support it. A register dominated by relationship holders does not supply that support.
• The board route was blocked from inside. Directors drawn substantially from within the company had limited independence from the management whose capital allocation was at issue.

Each mechanism was blocked by the same structure, which is why the condition was stable rather than merely persistent. A market inefficiency that is protected by an ownership structure is not an inefficiency in the usual sense — it is a stable equilibrium, and it persists until the structure changes.

This explains why the eventual catalyst was institutional rather than market-driven. No market participant could break the equilibrium, because the equilibrium was precisely what prevented market participants from acting. Only a change in the rules governing the structure — governance codes, disclosure requirements, exchange pressure, stewardship expectations for domestic institutions — could do it.

A gap that persists for decades in a sophisticated market is not being missed. It is being prevented from closing, and the useful question is what is doing the preventing.

The corollary for anyone assessing the reform's durability: the thesis depends on the structure continuing to unwind. Cross-shareholding levels are disclosed and are therefore the best single indicator — better than any statement of intent, because they change who votes.

What an allocator could act on

Compute the cash-adjusted return on operating assets. Strip idle cash and non-operating assets from equity and recompute the return. A company with a mediocre reported ROE and a large cash balance may have a good operating business, and the gap between the two numbers is the size of the opportunity. This is arithmetic from published financials.

Watch capital returned, not plans published. A comply-or-explain governance regime produces documents. The measurable test is buybacks and dividends actually paid, and the divergence between the two is the clearest evidence of whether the reform is real.

Track cross-shareholding unwinding. It changes who votes, which is the load-bearing element of the entire thesis. Disclosed by companies and aggregated by several sources.

Hedge or accept the currency exposure knowingly. A foreign investor's return in this market has been substantially determined by the yen rather than by equity returns. The 2022 regional report describes the divergence at its widest. This is the single largest determinant of a foreign investor's experience of Japan and it is frequently treated as an afterthought.

Recognise the return source as uncorrelated and value that separately. A case that does not require growth is driven by different variables from everything else in the region. In a portfolio otherwise exposed to trade, technology supply chains and the dollar cycle, that is genuine diversification — which is the strongest argument for holding it as a separate allocation rather than inside a regional one.

Accept the timescale. Governance change is slow. A thesis requiring years is difficult to hold through periods when it is not working, and the principal risk to it is not that it is wrong but that it is abandoned before it compounds.

What 2019 established for Japan

• The sub-book condition was shown to be a capital allocation problem, not an asset quality one.
• Idle cash depresses value twice — through ROE arithmetic and through a discount for permanence.
• Cross-shareholdings were the load-bearing structure preventing the normal discipline mechanisms from operating.
• The catalyst was institutional, which makes it broad, slow, durable and reversible.
• The return source does not require growth, making it genuinely uncorrelated with the region's other drivers.