Monday 12 August 2019

Why the Fed has to cut 100bps, but will only do so after a market crash


I have previously argued that QE repatriation, which I thought would be driven by QT, is a new Plaza Accord that would put the Yen and probably the Euro into 2-3 year structural bull markets.

It seems we are now there, via a different driver. With USD Libor above the yields of Treasuries and other fixed income securities, financial and other investors are incentivised/ forced to take profits and it seems repatriation is going on. This is also causing the Euro to now rally in risk off days and the USD to sell off. That's the opposite of prior relationships and shows just how much the US has been dependent on external funding in this cycle and on an ongoing basis, given Trump's deficits.

You can clearly see the repatriation in the Yen. The Yen bottomed in Q4 as US 10 year yields peaked and has tracked the US 10 year lower (stronger Yen). Recently the US 10 yield going below USD Libor has accelerated both the Yen rally and the fall in yields; they seem to be reinforcing themselves on a reflexive basis.



In my opinion the Yen could go to 65 before it has any real valuation problem, but I expect that will take several years. It has until recently been one of the cheapest majors on a REER basis.


I also think the JPYCNY rate could go over 7.5c and perhaps over 8.5c later, which would represent a full unwind of Abenomics and a much more competitive CNY.


According to the headline Fed figure, US banking system excess reserves are $1.37Tn.


However JP Morgan, BaML and others have argued that true excess liquidity is now tight and that amongst other factors a few banks hold a lot of excess reserves, while others are struggling to fund themselves, such as primary dealers who are holding what seems to be an ever growing pile of >$200bn of TSYs.



To finance which they have added about $125bn of net repo financing since Trump cut taxes last year.



Into this mix we see the US Treasury increasing Treasury issuance significantly over Q3 and Euro and Japanese investors liquidating fixed income.



In summary then, we have catalysts for USD liquidity to tighten a lot in Q3 and for US investors who buy the Treasuries being sold to have to sell other assets to fund those purchases. To rectify this dynamic JP Morgan and BaML conclude that the Fed will have to restart OMO/ QE.

However that doesnt resolve the fact that US Libor is higher than TSY yields. To fix that the Fed will need to cut about 100bps, assuming the curve doesn't fall further before the Fed cuts.

However the Fed/ Powell have said they will only cut rates on a lagging basis to data/ markets. So they wont pre-empt any move. Instead financial conditions will have to tighten significantly and the fastest way for that to happen is equities and credit crash into September.

It goes without saying that both Trump and Xi have the geopolitical equivalent of 'sell buttons' that can be pressed at will.

This whole dynamic of liquidation causing market falls and then Fed rate cuts is hardly positive for the USD bulls aginast the major currencies, albeit the USD is likely to rally against EM FX particularly as countries like Brazil slide into recession and South Africa gets downgraded. 






Wednesday 7 August 2019

Why is the Euro rallying on risk off?

Normally in a risk off the USD, TSYs would rally and equities and the Euro would fall. The Yen usually rallies due to capital being repatriated.  

But we have seen in early August that the Euro has rallied on risk off. That can only be flow and position unwind driven.




So which flows and positions are being unwound? Aside from all of the other financial assets like ABS, corp debt, equities, if we just look at Treasuries we might get an answer.

Europe, including the City of London have bought $2.8Tn in Treasuries this cycle, net, plus other US financial assets as part of QE leakage. China and Japan also bought but stopped accumulating TSYs in 2015 and have even cut back on holdings recently.




In this cycle the USD has needed $400-500bn a year to support the CA deficit and the US government has seen debt rise from approximately $10Tn to $22Tn.




So out of the $12Tn rise in US govt debt, these four countries have bought $3.5Tn net and the Fed bought $2.1Tn (from $2.46Tn peak). So QE accounts for $5.6Tn or 46% of the total.

Obviously the higher yields available in US government debt combined with the AAA rating made this attractive to European financials even net of xccy swap basis costs. Unhedged investors have also benefited from the strong USD since 2013.

Unhedged there is a large yield pick up for TSYs. However the flattening of the curve and the fall in yield differentials has made them unattractive or even negatively yielding after xccy swap basis is taken into account.



US 3m Libor is currently 221bps and Euribor is -38.9bps. So a 3 month FX hedge would cost 260bps plus the xccy basis which moves around and I don't have available.

However even if the basis is zero your US 10 yr would yield today 1.63% in USD and that would leave you with -97bps running plus any xccy spread in addition, plus 8 years of interest duration risk vs a 3 month FX hedge. In comparison the Bund is yielding -59bps running and can be repo financed at -47bps (a few months ago the Bund was positively yielding).


As such as risk off hits the markets and TSY yields fall, is it any wonder that European and Japanese investors whether they are FX hedged or not are taking profit and repatriating capital?

So all things being equal this means they sell US TSYs for USD cash and then use that cash to repay any funding/ credit associated with the position such as repo, which is a form of Quantitative Tightening. Similar to the US Treasury rebuilding its cash balances over the next few weeks.  

The USD is also very high on a REER basis causing Trump to loudly complain. Amusingly the Renminbi is the only major that looks more expensive, mainly down to the quasi USD peg. Mexican Peso looks cheap. The French Franc and Deutschmark on a Euro look through basis have also never really been cheaper.


So a Euro rally on risk off is somewhat counter intuitive and you might expect the Euro and Sterling to slide again into Brexit, which is more of a localised story. 

However you will still have this capital repatriation tendency in risk off. Put it another way, the Euro is a now a carry trade funder, same as the Yen.

Another corollary is the Fed will have to cut 100-125bps for 10yr TSYs to have positive carry swapped into Euros. Otherwise they risk forcing the sale of $3.5Tn of them.  










Monday 5 August 2019

HSBC's exposure to Hong Kong real estate

So conventional wisdom is that post-Basel III the banks hold a lot of capital against loans and are run conservatively. And in a normalised market this is very true I think.

However when you are calculating LTVs and RWAs and PDs against bubble valuation levels, are they still appropriate? If you calculated it against replacement costs, the LTVs would go through the roof, and so would RWAs and the banks would be left with an CET tier 1 equity deficit to be covered by a rights issue. Any losses and higher RWAs on impaired loans would further cost equity.

So Hong Kong real estate which yields 1-3% rental yields in many cases, vs a Prime lending rate which is 5.15% is an enormous, negatively carrying bubble, propped up by speculation and Chinese capital flows.

HSBC is the 800lb gorilla in a banking system where M3 is >5x GDP.

M3 and Household debt to GDP:







Here are some excerpts from HSBC's financials and Pillar 3 statements for Dec 18.


HSBCs 2018 AFS. $300bn plus exposure to HK and China.




Downside risk scenario involves small house price drops and ongoing economic growth.



Trade exacerbated downside drives a greater slowdown in 2019 only with recovery in H2-20.

 


$55bn of Category 3 loans to wholesale borrowers. Category 3 typically has a 100% RWA under Basel III.

 
$58bn of CRE, $2bn impaired

 
About $15bn in higher LTV wholesale loans:
 


$108bn in personal loans, of which about $3bn are impaired
 
 
About $25bn in loans above 51% LTV
 


However most of their book has RWA charges determined by their Internal Ratings Based approach:


Using their IRB approach retail mortgages have only a 0.7% probability of default and a few percent of loss given default. They calculate a 20% RWA charge against this.




Wholesale loan PDs are also under 70bps and seem to be applying a 30% RWA charge.




Only by IRB derived models and applying low PDs and low LDG rates have they been able to book these loans at 20 or 30% RWA levels and as such 'Category 1 loans'. But Category 3, 4 or 5 loans carry 100% or 150% RWA charges. So any large falls in property prices and rise in PDs will absorb a lot of CET1.




As a reminder after the 1997 bubble imploded the Hong Kong residential index dropped from 172 to 59 between 1997 and June 03, a 66% drop in nominal terms and more in real.
 


And remember the prime lending rate is 5.15% and property yields 1-3%. Well they have 3 year interest deferred ARM mortgages to resolve for that.... I wonder if any of HSBC's wholesale loans book is against such loans?





Friday 2 August 2019

The bullish case for Brexit and a new investment led economic cycle

Most forecasts for the impact of Brexit are negative. Since its never happened before the process seems to be to identify disruption factors, link them to an assumed GDP impact and conclude that in all scenarios Brexit leads to some level of a worse outcome.

While I accept its difficult for statistical analyses to take into account positive events that could happen in the future, the fact is we run a £100bn a year goods trade deficit principally with 4 or 5 countries (China and EU countries) and that means there is about 1 million or so high quality manufacturing jobs sitting abroad that should be here, leading to lower direct taxes, higher government deficits and further second round effects.

The UK (and US) post-1997 (the Asian mercantilist era) economic framework of gutted manufacturing and double deficits is simply not sustainable or optimal and we have seen the election of Trump and Brexit as the political mandate to change it.

Per the BoE, Inflation Report, the consumer, business investment and exports have all been subdued for several years in particular since the 2015 Brexit Referendum Act.
I don’t see a reason for the consumer to lever up.

Business investment in particular has really lagged, with companies preferring high margins on exports rather than increased export volume and that is one reason why the trade deficit hasn’t rebalanced much despite Sterling weakness since 2016. Per the chart below, business investment could bounce 40% or so vs. previous economic cycles.


The tightening of the labour market and rising wages is also causing companies to focus on using capital over labour to boost output, which again supports investment.



They have long term growth potential at 1.4% vs 2.9% pre-GFC. The two main dragging factors are capital and productivity. So a rebound in investment should boost growth potential.



I have long argued that MAGA/ Trumponomics involves: higher wages; forcing higher investment; leading to higher productivity; improved trade balances and higher interest rates/ discount rates over time, plus zombie companies going bust. 

Also that fiscal deficits and low real rates can be used to transition from the credit led economic cycles seen since the early 1980s to a new wage/ inflation/ investment cycle, without having a meaningful recession in the middle

So UK government fiscal stimulus combined with the Brexit windfall discussed above, would lead to a pick-up in investment, productivity, exports, wage growth, inflation etc.

If that happens, the MPC expect rate hikes to be likely over the next two years, but for rates to still be much lower than nominal GDP.
(For the sake of clarity this article discusses what may happen in 2020-2022, not in the rest of 2019, as I think the economy will deteriorate into Sept/ Oct. I also think a base rate cut to 50bps is quite likely if we have a hard Brexit at the end of October. ).