Friday 31 January 2020

Brazil, now the 'Bolsonaro bounce' has worn off

I previously had a look at Brazil, given the excitement when Bolsonaro was elected. There was a great deal of hope in soft indicators like sentiment and the Bovespa welcomed him being fairly resilient on the Q4-18 sell off and doing well in the first part of 2019.

I wrote:

"Im guessing a bit more yield compression first though so curve flattens first and equities rally on hope. With investor flows into EM expected in H119 there is scope for the curve to bull flatten into early next year. The BOVESPA on a mid-teen P/E is breaking out by the looks of it"

So how did things go?

The stock market made a high in July/ August in USD and BRL, but BRL weakness later in the year put it behind the S&P into year end and YTD in 2020 as of writing it is weak in USD, in BRL the Bovespa peaked in January. 

The index is cyclical, mostly banks and resource companies, so it did well to keep pace with the S&P last year I think.

16x P/E, 2.6% div yield, 2.3x book. Given the index is cyclical its a fairly middle of the road valuation. 

In fixed income the curve bull flattened as expected. Initially after the election then again during last year:

There was a little over 230bps compression in the 10 year, but bear in mind the 10 year peaked at over 16% yield in 2016...

Coming into this year front end rates are down to 4.5% as of December!

Which is as low as they have been in this Century. 

The hopium didnt translate into much in the real economy and the recovery has been fairly tepid and inflation actually started to rise last year as the currency weakened again.

GDP Growth YoY:

 Annual GDP growth:


Unemployment rate:

The sentiment indicators slackened off over the summer as some fears of a recession picked up but Business confidence is on a high into year end, perhaps helped by rate cuts? Consumer confidence is below 50 though.

Services PMI:
Manufacturing PMI:

Business confidence:
 Consumer confidence:

Brazil's exports:


Iron ore and oil products are key exports and oil prices have been dropping, iron ore has held up so far. But demand from China could shrink in Q1 if the virus scare drags on. 

Currency at 4.20/ USD is middle of the road on a REER basis, would go lower on commodity weakness or higher on EM growth rebound later in the year. The low front end rates have reduced carry attraction and caused some corporates to borrow in BRL and pay off hard currency loans which put some downwards pressure on the currency last year. Portfolio outflows on risk off and the poor carry differential could sent it quite a bit lower.

The private sector has been deleveraging since 2016 with some growth in household debt offsetting a fall in business lending.
Household debt/ GDP:

Private sector lending:

So in many respects, following a 2016 downturn, Brazil looks more or less mid-cycle. Weak commodity prices and private sector deleveraging are a drag on the economy, but you also have recovery dynamics and a pick up in business confidence.


Below are the IMF forecasts. I would pick out a few things. There are some oddities like an investment led recovery, consumption remains weak, public investment remains low, unemployment only slowly declines, FDI shrinks in USD and % of GDP terms, the government shrinks the primary deficit over the next two years but is still borrowing over 7%, mainly due to interest on gross government debt which is heading towards 100% of GDP in BRL mostly.

There is not much in there for the consumer leading growth or a global slowdown either.

The IMF want them to invest much more in infrastructure, which would help open the economy to more trade and engage in other reforms in taxes, fiscal consolidation, pension and public sector wages reform. 

The government is bloated:

Yet doesnt manage to spend the money on investment, instead spends it on high public sector wages and pensions, leaving little room for anything else. 

The pension reform bill is expected to deliver 2.5-3% of GDP savings by 2030, but doesnt look like it goes far enough. 

Privatisations could free up $75bn or so, but that is not much on a $1.8Tn economy.

Lack of infrastructure investment:

The poor transport infrastructure, combined with size and geographical remoteness plus anti-trade barriers reduce trade potential:

IMF DSA stress test:

Its not much of a stress test, for example rates were 14% 4 years ago and there is only a small fall in GDP and no inflation spike; surprising since the BRL normally tanks 50% in a recession. 

Nevertheless even this mild stress test puts government debt as high as 120% by 2022.

So overall, just another poorly led emerging market where the government has spent more than it collects on taxes for years and offers state employees gilt edged pay and pension packages while the private sector struggles with low wages, corruption and poor infrastructure. 

They are borrowing to roll up interest on the debt. But the debt is internal, so its more about internal distribution than external imbalances. 

The country has a history of inflation and currency weakness, so a middle of the road REER valuation and just 4.5% SELIC rate is not very attractive even when markets are stable, and in fact the BRL weakened meaningfully H2 last year. 

Writing this in late Jan, it looks like the US is about to lead us into a risk off phase followed by more stimulus. So I wouldnt be surprised to see BRL pushing in the direction of 5/ USD in a risk off, but after that in many respects the country looks mid-cycle. 

Wednesday 29 January 2020

The Fed inverts the curve deliberately?

The FOMC was supposed to be a snoozefest. 
But the FOMC hiked IOER to 1.6%, which is more yield than most of the o/s Treasury stock, which the primary dealers have to buy if you want to sell... Err, what?!?
Isnt this what happened in Sept/ Oct? - which forced 3 rate cuts and not-QE4 to avoid a crash?

On the other hand, assuming they know what they are doing, maybe these Dems at the Fed have it all worked out. Keep the plates spinning for just a bit longer, then blame Trump in Q2.
So I think the T-bonds are going to rally from here until the Fed panics, cuts and restarts QE/ twist and then Bill Gross et al sell the bonds to the Fed.

Saturday 25 January 2020

How close is Sri Lanka to a debt crisis?

So I got around to reading the IMF's last Article 4 report on Sri Lanka. These are the findings.

Ever since Sri Lanka restructured the Hanbantota port and airport, foreign bond investors have been selling down Sri Lankan bonds.

They have sold about 60% of what they previously own. There is therefore little/ no market at this stage for Sri Lankan govt bonds to foreign investors.

External debt of 64% of GDP and likely to rise, that is at a level where if there is a down turn, then the IMF would impose strict fiscal reforms on Sri Lanka.

Out of current expenditure of 2.5Tn LKR, 1.6Tn is on interest and subsidies alone. Effectively they are borrowing money to cover transfers, not spending in the economy or investing in capacity.

If the oil price falls, which I think it will, remittances from the GCC might slow down and the country is dependent on them. They also are dependent on FDI and issuing $1.3bn in net new foreign bonds.

High levels of gross financing as well across several areas.

Sri Lanka is a net debtor country. But instead of borrowing from abroad to invest in productive capacity, I think much of the borrowing has been to finance consumption, roll up debt and subsidise loss making SoEs. As such, the more the borrow, the deeper the financial hole they are in.

Debt has grown 28% and GDP about 20%, so debt is growing almost 1.5x as fast as GDP.

The IMF's DSA assumes Sri Lanka increases its primary surplus by 2.8% to +1.3% when it has averaged -1.5%. That is quite a large consolidation. It also assumes no real FX devaluation and real economic growth goes up to 5%. Is that growth rebound realistic in the face of a large primary balance adjustment?

It would be easier to just cut local interest rates and use macro prudential tools to stop bank credit growth accelerating.

The IMF's 'combined macro shock' scenario is not that much of a shock. It still has positive economic growth and a positive primary balance. A proper recession is likely to look much worse than this however. Their combined shock scenario has debt peaking at 105% of GDP. Way above the level that is sustainable and would effectively mean a debt restructuring would happen. In a typical EM FX crisis it is more likely to see -10% GDP growth.

Below is simply a false statement made by the Finance Minister. Sri Lanka's tax level is not out of the ordinary for a country with its level of GDP. Its similar to Egypt for example.

The issue is a large part of the spending in on debt service.

They think the answer is to tax people earning money from their work? Instead you should tax the people who own the debt. The easiest way to do that is cut interest rates and devalue the currency, while encouraging private sector investment and work, not taxing it more.

The SoEs should present a plan to earn a profit and justify their continued public ownership. If they cant do that they should be sold to the private sector.

My understanding of this is that a 2% primary surplus is normally not achievable over the medium term for most large countries.

For example, Jamaica as a small country with remittances of over 20% of GDP was able to run a 7% primary surplus.While in Greece, getting to a 1.5% primary surplus caused a depression.

For a small country like Sri Lanka it might be possible but only because a deficit is run elsewhere,  aggregate demand is held up and growth continues.

Basically for it to not slow growth the reduction in demand in the economy from the government needs to be replaced by demand elsewhere. The two easiest ways to achieve that is a devaluation which makes foreign remittances more valuable as a % of GDP, policies that force bond holders when they get repaid to invest their money into the real economy (such as low interest rates) and an improvement in the import/ export mix via higher service exports and more domestic goods production. Another way is to slow spending growth at a rate less than nominal GDP growth, so it rebalanced over time.

Put simply, engaging in a front loaded fiscal consolidation when you already have high debt and a slowed economy, without counterbalancing supply side measures, is likely to fail and lead to higher debt/ GDP later. 

So what policies does the government have to rebalance the economy, while maintaining grow, while at the same time the government hikes taxes? If the answer is they don't, then the government's fiscal consolidation will just slow growth and cause them to miss their DST targets.

You have to wonder how much shit will be unearthed by making 200 SoEs publish financial statements. It is frankly incredible they have been allowed not to. They should also be investigated for fraud.

While in the long run (10+ yrs) this might be great, opening the doors to Chinese and Indian volume producers versus inefficient Sri Lankan producers might see job losses/ bankruptcies in the short term. Sri Lanka as a tiny economy will always struggle to have volume producers in many goods markets.

The new government's manifesto can be found here. I couldn't see any big ideas in there, just marginal improvements on what is already happening. It reads like it was drawn up be a mid-level staffer. For example, two key sectors, apparel, which is $5bn of exports, got three paragraphs and the IT sector, one page. Even though IT services is one of the areas they have a good chance of growing in. I have pointed out to them, if they want to get to the next level, they need to create globally competitive jobs that pay $15-50k and areas like IT or health care are the two most likely areas.

So in conclusion, I think the country is just a few missteps away from heading towards a debt restructuring and a much tougher IMF programme.

Friday 24 January 2020

The US curve is now inverted out to 7 years

The US curve is only 15bps from inversion again and the 10year has completely given up the so-called Q4 economic bounce. 

The US curve is now inverted out to 7 years. 

The bond market thinks the Fed needs to cut at least 2 or 3 more times. 

To be fair to Powell they acted pretty quickly in the Autumn as without that I would have expected a crash