The Macro Man Team put up an excellent post this weekend on the reflation case for Japan. It is highly recommended reading, and does a very good job of elucidating the bull case for Japan, and bear case for the Yen.
Regular readers are probably aware that I’ve disagreed with many of the reflation views, (and has been wrong thus far) so I thought this is a good time to re-evaluate that view as well as rebut some of the points in the post above. Given the multifaceted points that will be discussed, I will attempt to summarize each point made in a sentence, and follow that with a rebuttal. As always, counterpoints are more than welcome.
Basic Balance of Payments correlation with Yen is limited due to corporate FX hedging. The improvement in BBoP in 2004-2007 coincided with a period of Yen weakness, while the drop in BBoP from 2007 onwards coincided with Yen strength. Just as the improvement in BBoP didn’t have a big, measurable impact on the level of FX, its deterioration may not as well. In fact, if historical trends correlations hold, this could actually a Yen positive since it could require additional repatriation from abroad.
Hypothesis: The 2005 experience suggests asset prices can move substantially further.
It is unclear if 2005 a good comparable. MM thinks it is the end of deleveraging that drove the recovery in USDJPY. If that is the case, then the backdrop does not appear similar then vs now, as the Japanese private sector is not coming off of the end of 15 years of deleveraging. Also, in 2005, USDJPY increased ~17% off the lows in an 11 month period. The current move is already +15% over a similar time frame.
Furthermore, the Nikkei was trading at a TRAILING PE multiple of 16x at the end of 2004, and the world was just a year out of a global recession. On a FORWARD basis, it is trading at 17x now, the highest level since before the earthquake, with a much weaker global growth backdrop. Valuations could certainly go higher (if it goes to the 2006 high of 28x, that’s a 65%+ move from here) – but the idea is that a further move here necessitates both an estimate adjustment as well as a momentum chase.
Hypothesis: "Can policymakers raise Japanese inflation expectations towards 2%? … evidence from the US & UK is that if central banks are aggressive enough, they can move inflation expectations to where they want them"
The assumption there is that central bank’s ability to influence inflation expectations is stable over time, and can be maintained beyond historical averages. The evidence does not fully support this. The Fed and BoE have increased their balance sheets at a much more aggressive rate than the BoJ, but Inflation Breakevens (US in white, UK in orange)) have not moved out of long term ranges. The BoJ has been expanding its balance sheet for a decade, and Japanese breakevens are now already at the highs.
Hypothesis: The scenario for the evolution of real rates suggests further USDJPY upside
My biggest issue there is probably the implicit assumption that real rates spreads were the main driver of USDJPY over the past 9 years. The two series really only correlated since 2007, as global deleveraging took hold. Also, they only correlated in one direction. This is typically a warning flag, because lots of indicators appear to be correlated simply because they move in the same direction. I.e. the lemming population and returns from a short vol strategy. The risk is that the relationship suddenly breaks down. A chart of nominal rate spreads vs the cross suggests USDJPY is fairly valued here, and a widening of the rate differential of 150bps could only drive a move to 95, assuming PPP stays constant!
Hypothesis: Long dated JGB yields could move a lot higher.
The argument there assumes that the BoJ’s can sustainably increase long term expectations. I have already noted that the US and UK experiences does not suggest that this is certain. I will also reiterate that QE induced inflation requires a monetary transmission mechanism. In Japan, only the public sector has the capacity to do so – so we will need to see a move toward growing budget deficits before the BoJ’s QE programs can increase inflation. With the budget balance already at -10% of GDP, it is not clear of there is sufficient political will go even further. Abe has already resigned as PM once, and his approval rating is not high.
Hypothesis: Nikkei EPS growth will surprise to the upside based on model outputs.
I wasn’t able to replicate TMM’s model because I wasn’t sure how it handled negative EPS prints. (which occurred in 2002 & 2009) When I tried to adjust for that, I get a negative coefficient for the core CPI variable. So let’s just look at the TMM’s model as is. Note that in TMM’s model, it appears that the EPS model was too high in much of recent history. In other words, the model under-predicted EPS in the 1990’s, and over predicted in 2000’s, which suggests that the model has been losing forecasting power. Also, IMHO, a 12.2x multiple using a model that projects substantially higher earnings than consensus is not particularly cheap vs an S&P 500 index that trades at 12.9x using forward consensus estimates.
I also tried to create an EPS model for Japan, with limited success. Only the ISM and a change in USDJPY had statistically significant T statistics and sensible signs. That model projects an EPS change of 20, implying a terminal 2013 value of 40, with a standard error of 15. (Assuming ISM of 53 and USDJPY @ 95) This compares to consensus estimates of 29, so that also suggests some upside, if not a substantial one, and an implied valuation that is richer than the S&P 500. Note that since my model uses the change in USDJPY, EPS increases resulting from this is not sustainably repeatable.
To conclude, my view is simply that both USDJPY and JGB yields are likely to be stable in this new range for some time, UNLESS the Japanese government commits to sustained increases in the budget deficit.
- A substantial increase in BoJ purchases of JGB’s is likely to more than offset private sector sales, and thus contain increases in nominal yields. If the budget deficit next year is 10% of GDP, that implies ~50trn Yen of net issuance. If the BoJ buys 5Trn a month, it will buy 120% of net issuance. (Note also that a sharp increase in yields will devastate Japanese Financials balance sheets…) Even if the BoJ is successful at increasing inflation expectations, it will probably be in the same manner the Fed was successful – Inflation Breakevens increase because real yields turn negative. Note again that if the government does not run worsening budget deficits, it will be difficult for Japan to get much in the way of realized inflation, which means the additional increases in inflation expectations is likely to be capped.
- The level of USDJPY is, IMO, predominantly a view of the victor in the battle between policy makers and secular economic forces. While I do think if the Japanese government is willing to behave like the WeimarRepublic, they will win in the short run, my view is still that many of the macro forces that drove the currency to these levels continue to persist. With the cross already near the ~90 level that Abe has declared as ‘fair,’ it is not clear how much farther the cross will move. IMO, the market is already pricing in a 40% higher increase in Yen money vs USD money. I think the BoJ will need to do 10trn Yen of QE per month to get there, equality a 75% increase in its balance sheet over 12 months. Anecdotal reports suggest expectations are for half of that.
- With real growth unchanged, Nikkei EPS growth is likely dependent on how far the Yen moves, and is not likely to improve for more than a year. Current valuations do not suggest a bubble, but given the long term dynamics, current valuations do NOT support an overweight recommendation.
- IF the Japanese government proceeds with an expanding budget deficit, and IF the electorate consents, then it will be time to purse WeimarRepublic type trades. Long Yen denominated assets, FX hedged, and short Yen positions will all make money.
Thanks again to TMM for their post. As always, comments are welcome.
- BIS put out revised liquidity standards for banks that the market views as more flexible than anticipated. The release gave banks 4 more years to meet international liquidity requirements.
- The Journal posted an article about the upcoming stress test results for the 19 largest US institutions. The buyback request from Citi is expected to be “minimal.” JPM is expected to submit a plan that calls for both a dividend increase and a share repurchase. Morgan Stanley won’t focus on a buyback or dividend, but instead on winning Fed approval to take full ownership of the Smith Barney unit.
- Sentix investor confidence in the EU increased to -7 from -16.8, above estimates of -14.2.
- The Japanese government will announce around 12 trillion yen ($136 billion) in fiscal stimulus measures to boost the nation’s shrinking economy, Japanese media reported today.
- Melbourne Home Sales hit 16 year lows: Macrobusiness
- NYT: In 2012, the index never had a daily close below the closing level for 2011. Going back to 1928, there have only been eight other years where the index went an entire year of trading up year-to-date every single day. The last time was in 1979.
- Mon: Australia Trade Balance
- Tue: EU Economic Confidence
- Wed: Canada Housing Starts