A short study of the recent Japanese crisis
Based on public data sources providing Japanese stock returns and bond interest rates, this article provides a quantitative analysis of the trajectory a Japanese passive investor could have experienced with his portfolio during the 1980-2016 time period. Japan suffered from what was probably the most severe stock crisis in modern history, starting by the end of 1989, taking nearly two decades to finally somewhat recover, and then being slammed again by the financial crisis of 2008. There is no equivalent in U.S. history of such long lasting stock crisis, and as such, it is a sobering example of what could happen.
(A follow-up article to this study was published here, exploring various questions raised by readers).
Japan was extremely successful during the 70s and early 80s, gaining impressive ground on a worldwide basis. The Japanese crisis at the end of the 80s had a rather dramatic effect though, as shown by the following graph from the Credit Suisse Investment Yearbook 2013. It is worth noting that Japan market weight was larger than the U.S. for a brief period of time, following twenty years of expansion.
The Japanese central bank (Bank of Japan) tried to manage the crisis reducing interest rates over time but the troubles persisted to this day. Late 2015, Bank of Japan took the extreme step of moving to negative interest rates. Here is the average of 1yr to 10yr interest rates during the 80s until now (assembled by the author, see more below).
The decades-long economic troubles impacted the exchange rates as well. Here is a graph from FRED showing the evolution of the Japanese Yen to US dollar exchange rate during the same time period. Note the sudden appreciation of the Yen against the US dollar occurring right before the stock crisis of 1989/90 hit Japan.
Finally, the inflation in Japan dropped to very low levels, with periods of deflation, as illustrated by this graph from FRED. In 2015 and 2016, the inflation rate remained very low (0.10 and 0.30 respectively when computing Dec to Dec; see here for a full history since the 50s).
A detailed description of the Japanese asset bubble of the 80s, subsequent burst and aftermath can be found on Wikipedia.
Gathering stock and bonds returns; first-level analysis
This article was made possible by the refreshing openness of the Ministry of Japan (for interest rates) and Nomura Securities (for stocks indices).
The Ministry of Japan provides a full table of interest rates for various durations (click here). Based on such information, AlohaJoe was kind enough to parameterize his bond fund handy software (described here), and derived monthly returns for maturities ranging between 4 and 10 years, hence approximating an Intermediate Term government bond fund. The author ran a quick sanity check against a Barclays index tracking government Japanese bonds at large (Morningstar chart here), and the trajectory in overlapping years was similar enough to convince us that the 10-4 simulated bond fund provides a solid basis. As a side note, the Vanguard fund JPY tracks the same Barclays index.
Nomura Securities provides various stock indices about the Japan market since 1980, in collaboration with Russell. This not only documents the overall market, but also the classic splits using size and value/growth factors. Monthly and daily returns (including dividends) can be downloaded here, providing a rich base of information.
Thanks to those various data sources, the following chart (expressed in nominal terms) was assembled by the author, a simple growth chart of one Yen invested on Jan 1st, 1980 till the end of 2016. This doesn’t factor in expense ratios, just the trajectory of the indices, and as such is somewhat idealized. It is unclear if any form of low-cost index fund was actually available in Japan in the 80s and 90s.
Various things quickly jump to the eye:
- Inflation was indeed very mild (less than 1% a year over the time period).
- The (domestic) total market skyrocketed in the 80s, then fell very hard (more than 50% from the peak in Dec-89), stayed low for 15 years, seemed close to recovery in the mid 2000, then fell again hard (50% again) with the financial crisis in 2008, and is only catching up by now (hopefully) some 30 years later. That is a heck of a drawdown, to say the least.
- Small stocks did not seem to make much difference with the total market (except possibly in the last few years).
- Small Value stocks displayed a pattern somewhat similar to the US, with occasional sudden gains against total market, long-lasting drops or sideway trajectory, and making quite a difference at the end if one stuck to the course for decades (requiring nerves of steel!). Note that the Mid Value display was disabled from the chart for clarity, its trajectory being very similar to Small Value.
- Finally, bonds displayed a very steady trajectory during the whole time, providing welcome stability and actually better returns than most stocks (except small value).
Studying a few asset allocations
Most investors around the world learned the value of diversification. Some US investors may be less sensitive to the virtues of such diversification due to the sheer size of the US economy and its excellent track record, but passive investors located in any other country should be keenly aware of the risks involved in solely investing with domestic stocks. This is easy to say in hindsight though and wasn’t necessarily properly acknowledged some 30 years ago, notably when Japan was on a meteoric rise. Let’s compare fully domestic asset allocations against worldwide allocations treating Japan as a strong tilt.
In order to do so, let’s rely on MSCI, which provides a long history of worldwide returns with the MSCI World index. The corresponding (non resident) data series on Morningstar can be found here. This series is in USD though and needs to be converted in JPY (using the exchange rates obtained from FRED). This conversion actually hurt returns given the trajectory of exchange rates during this time period, which added insult to injury for Japanese investors. Let’s assume that an index fund tracking MSCI World would have been available to Japanese investors while again neglecting any expense ratio to simplify.
The following charts are growth charts for an initial investment of one million Yen in 1980, and no contribution nor withdrawal afterwards. Those comparison charts are inflation-adjusted (not that it makes such a big difference). The first asset allocation (AA) is a simple 60/40 allocation, all Japanese. The second asset allocation splits the stocks, 40% World and 20% Japan to add a domestic tilt (somewhat arbitrarily; also remember that Japan is a significant part of the world index, hence is counted twice here).
For sure, the diversified investor might have been frustrated until 1989, but then the trajectory speaks for itself. Drawdowns were painful, but limited to 5 years or so when diversifying, while the Japan-only investor would have waited some 15 years for a brief recovery before the financial crisis hit. Bonds did help a lot smoothing the trajectory, thanks to decreasing interest rates.
Now let’s assume that a more bullish investor decided to go 80/20 in the 80s. First AA remains domestic only, and let’s add a Small Value (SV) tilt, e.g. Total-Market 70%, SV 10%, Bonds 20%. While the diversified investor could have done something like World 50%,Total-Market 20%, SV 10%, Bonds 20%. This would have made for quite the roller coaster, and it stretches belief that the domestic-only investor would not have changed course at some point. Note that the 2008/09 financial crisis would have hit both investors especially hard.
It would be very interesting to extend those charts by ten more years though. Solid exposure to bonds would certainly have helped those investors during the troubled decades while interest rates were decreasing, but the consequences of going to near-zero or even negative rates in recent years are very unlikely to be bonds-friendly in the near future. Also, will Japanese stocks finally truly benefit from a long-awaited period of growth remains to be seen.
As a side note, an additional test performed by the author involved a 10% allocation to Gold (price, not miners) converted in Japanese Yen. Unsurprisingly, this didn’t help.
Retirees and withdrawals
Now let’s add withdrawals to the trajectory. To keep things simple (although not quite realistic), let’s use the good (?) old constant withdrawal scheme (fixed % of the starting portfolio, inflation-adjusted over time).
Clearly, the “4% rule” would not have worked well. Even with a conservative 60/40 (fully domestic) asset allocation, the portfolio would not have survived 30 years. When starting by the end of 1989, one has to go down to a 3.3% fixed withdrawal to keep the portfolio barely in the black over such retirement period.
This isn’t really a fair point though. Between Jan 1980 and Dec 1989, Japanese stocks (total market) increased by more than 650%, and a 60/40 allocation increased by more than 350%. The high point reached by the end of 1989 just defied gravity. It makes little sense to solely focus on the fall while ignoring the huge bull market that preceded it. This being said, even retiring at the end of 1988 (one year before the extreme peak) would have seriously challenged a 3.5% constant withdrawal.
A more diversified portfolio (back to our 40% World, 20% Japan stocks, 40% bonds allocation) would have navigated the situation much more cleanly. Here is a graph comparing the domestic-only and the internationally-diversified 60/40 allocation, starting by the end of 1988, with a 4% constant withdrawal at the beginning of each year (the domestic-only scenario falling apart when the financial crisis struck; the diversified scenario navigating the consecutive crisis fairly well, all things considered).
Now what about a more aggressive portfolio? Quite obviously, the 80/20 domestic portfolio would have failed a retiree in a dramatic way, with or without a Small Value tilt.
Let’s compare instead the two internationally diversified allocations we previously defined, both with a strong world exposure, plus a significant domestic tilt. The first allocation (aggressive) including 20% bonds, plus a 10% Small Value tilt. The second allocation (conservative) being the same as the previous chart, with 40% bonds. Between the incredible beating Japanese stocks suffered from for decades and the painful trajectory of the currency exchange rate, it is interesting to see that the 80/20 allocation would have actually held its ground reasonably well. It would actually have done better than the 60/40 allocation for all starting years except 1988 to 1991, while doing reasonably ok starting from any of those painful years, except for the end of 1989.
Overall, it would of course have been much wiser to use a variable withdrawal method (e.g. Bogleheads VPW, Guyton-Klinger decision rules, etc) in such difficult times, but the key conclusions would remain the same.
Japan went through a stock market crisis of spectacular proportions which lasted for nearly three decades, and it isn’t clear if full recovery is truly in sight at the time of writing. A purely domestic investor would have been wise to buffer his/her stock investment with a solid allocation to bonds.
The wiser move though would have been to diversify between domestic stocks and international stocks (and optionally to use a small tilt towards Small Value). Under such circumstances, a conservative (60/40) or a fairly aggressive (80/20) diversified stock exposure would have allowed to navigate the crisis much better, even when retiring near the height of the bubble while using a naive constant withdrawal method.
A US investor perceiving that the US is ‘different’, being a larger and more diversified economy than Japan, should consider that Japan had (admittedly briefly) a larger market capitalization than the US before the crisis struck (see the first graph in this article). No economic power is too big to fail. Individual investors should probably put such home bias perceptions aside and simply hedge their bets by diversifying.