Warren Buffett sensed value in Japan’s legacy conglomerates that others don’t. Photo: AFP/Johannes Eisele

The world of money managers is abuzz over Warren Buffett’s Berkshire Hathaway’s purchase of approximately 5% each of five Japanese trading houses, stakes worth in total of about US$6.2 billion. 

Berkshire Hathaway reportedly may seek to increase its stakes in the five trading houses – Mitsubishi Corp, Mitsui & Co, Sumitomo Corp, Itochu and Marubeni – to as much as 9.9%. Concurrently, Berkshire Hathaway issued yen bonds for the first time.

The 430 billion yen worth of yen bonds included 5-, 7-, 10-, 15-, 20- and 30-year-terms. 

The biggest offering is the 10-year debt at 146.5 billion yen carrying a coupon of 0.44% versus recent 10-year JGB close of 0.04%. So the risk premium on AA-rated Berkshire is 40 basis points above the Japanese sovereign.

“The five major trading companies have many joint ventures throughout the world and are likely to have more,” Buffett said. “I hope that in the future there may be opportunities of mutual benefit.”

The impression Buffett gave was that he likes Japan for the long term because it can open up a whole new world of investment opportunities. Meanwhile, Berkshire Hathaway has been reducing its exposure to North American assets 

For Berkshire Hathaway, with a cash hoard of $146.5 billion, the $6.2 billion in five trading houses – sogo shosha in Japanese – is peanuts. The investments are held by Berkshire’s insurance company, National Indemnity Co.

Some have speculated that the move is in line with Berkshire Hathaway’s recent move to increase LNG assets as the price of liquefied natural gas has increased 40% in recent months. In July, it acquired the gas transmission and storage business of Dominion Energy.

The trading houses all have LNG assets around the world. Indeed, Mitsubishi was the first in the world to transport LNG from Alaska to Tokyo back in the 1960s.

Low valuation investment?

Others have speculated that this is just another one of Buffett’s acquisitions of stocks with low share price-to-book value ratios, also known as low PBR stocks. Is that a good valuation measurement for trading houses?

This writer was once an analyst covering Japanese trading houses. My research suggested that their share price was very sensitive to return-on-equity, or ROE. The relationship can be shown this way: PBR = ROE x PER.

Generally speaking, ROE of 10% suggested PBR of 1.0x. When ROE rose to 20%, PBR approached 2.0x. And when ROE dropped below 10%, PBR dropped below 1.0x

In order for this valuation formula to work, the PER, or the price-earnings ratio, of the trading houses must be ignored. And indeed their PER goes all over the place; the market does not show the usual bias against high PER or favor toward low PER.

Mitsubishi Corp’s head office in Tokyo, Japan, August 2, 2017. Photo: AFP

Current ROE outlooks on these companies suggest that the risk of reduction in PBR may be limited. That is another way of saying downside share price risk has already been priced in.

Peter Drucker of corporate analysis fame once said of these companies: “They are not traders; they are finance companies.”

People working at the trading houses disagreed vehemently: “There’s dirt under our nails; we get into the trenches; we are not some clean bankers.”

But their business model was essentially financing and the markets valued them as companies whose share price rose when the return on their assets rose and dropped when the returns dropped.

Quarterly earnings tended to be volatile because of assets being sold and assets being written off. The market was much more interested in the long-term returns of the assets they acquired, nurtured and held.

Simple spread play?

Some outsiders have wondered whether the Japanese might resent Buffett’s grabbing up the country’s “crown jewels.” Actually, no trading house is a crown jewel, although there have been times when it would have been tempting to call Marubeni a clown jewel; when I was researching, they were always making lousy calls and writing off a lot of assets.

I think this is simply a case of Buffett looking for a low risk-spread income for the time being.

Of the five targetted trading houses, Mitsubishi and Itochu have done the best in improving the returns on their assets and should increase their dividend payouts for the fifth year in a row.

Their dividend yields are rich: Mitsubishi’s yield is 5.17%, Sumitomo’s 5.00%, Mitsui’s 4.09% and Itochu’s 3.18%. Compare those with the 10-year JGB yield of 0.040% and Berkshire Hathaway’s newly issued 10-year coupon of 0.440%.

Berkshire Hathaway gets a spread of 4.73% on Mitsubishi shares and a 4.56% spread on Sumitomo shares at a time when 10-year US Treasury offers only 0.705%.

If you think the down-side risk on trading house ROE has been priced in and the risk of yen depreciation is minimal, what should you do with a couple of billion dollars of hoarded cash? Try a simple preleveraged 4.73% spread play.

A retired Tokyo-based analyst for a major US investment bank, Matt Aizawa now crunches numbers beside a lake north of the city.