19/9/16: Survivorship Bias Primers

Not formally a case study, but worth flagging early on, especially in the context of our Business Statistics MBAG 8541A course discussion of the weighted averages and stock markets indices.

A major problem with historical data is the presence of survivorship bias that distorts historical averages, unless we weigh companies entering the index by volumes. Even then, there are some issues arising.

Here are few links providing a primer on survivorship bias (we will discuss this in more depth on the class, of course):

• "Morningstar Takes a Closer Look at Survivorship Bias" https://www.ifa.com/articles/morningstar_takes_closer_look_survivorship_bias/
• "SURVIVOR BIAS AND IMPROPER MEASUREMENT HOW THE MUTUAL FUND INDUSTRY INFLATES ACTIVELY MANAGED FUND PERFORMANCE: http://www.etf.com/docs/sbiasstudy.pdf
• "When It Comes to Fund Performance, History Is Often Written by the Winners" http://www.wsj.com/articles/SB10000872396390444025204577545460615317218
• An example from actual data on construction sector activity: http://trueeconomics.blogspot.com/2010/08/economics-9810-irelands-construction.html

19/9/16: Big Data Biases?

A very interesting, and broad (compared to our more statistics-specific discussions in MBAG 8541A) topic is touched in this book: http://time.com/4477557/big-data-biases/. The basic point is: data analytics (from basic descriptive statistics, to inferential statistics, to econometrics and bid data analysis) is subject to all the normal human biases the analysts might possess. The problem, however, is that big data now leads the charge toward behaviour and choice automation.

The book review focuses on ethical dimensions of this problem. There is also a remedial cost dimension - with automated behaviour based on biased algorithms, corrective action cannot take place ex ante automated choice, but only either ex ante analysis (via restricting algorithms) or ex post the algorithm-enabled action takes place. Which, of course, magnifies the costs associated with controlling for biases.

One way or the other - the concept of biased algorithmic models certainly presents some food for thought!

19/9/16: US Median Income Statistics: Losing One's Head in Cheerful Releases

In our Business Statistics MBAG 8541A course, we have been discussing one of the key aspects of descriptive statistics reporting encountered in media, business and official releases: the role that multiple statistics reported on the same subject can have in driving false or misleading interpretation of the underlying environment.

While publishing various metrics for similar / adjoining data is generally fine (in fact, it is a better practice for statistical releases), it is down to the media and analysts to choose which numbers are better suited to describe specific phenomena.

In a recent case, reporting of a range of metrics for U.S. median incomes for 2015 has produced quite a confusion.

Here are some links that explain:

• http://davidstockmanscontracorner.com/ analysis of the data: http://davidstockmanscontracorner.com/deconstructing-median-income-bs/
• Charles Hugh Sith's analysis of the same set from a similar angle: http://charleshughsmith.blogspot.com/2016/09/fun-with-fake-statistics-5-increase-in.html
• Confirming the above, WallStreetExaminer.com: http://wallstreetexaminer.com/2016/09/chart-day-5-2-household-income-gain-vs-reality/ and a the same with links to original release of data: http://wallstreetexaminer.com/2016/09/median-household-income-rose-5-2-yeah-thats-ticket/
• The above are contrasted by the NYTimes coverage: http://www.nytimes.com/2016/09/14/business/economy/us-census-household-income-poverty-wealth-2015.html?_r=0 and Bloomberg view: http://www.bloomberg.com/news/articles/2016-09-13/u-s-poverty-rate-falls-to-post-recession-low-as-incomes-gain

So, as noted on many occasions in our class: if you torture data long enough, it will confess... but as with all forced confessions, the answer you get will bear no robust connection to reality...

19/9/16: FocusEconomics: The Italian Dilemma

Good post from FocusEconomics on the saga of Italian banking crisis: http://www.focus-economics.com/blog/posts/the-italian-dilemma-weak-banks-pose-risk-to-already-faltering-domestic-demand.

And an infographic from the same on the scale of the Italian banking woes:

Click to enlarge

It is worth noting that in the Italian banking case, asset quality crisis (NPLs etc) and compressed bonds returns (yield-related income decline due to ECB QE) are coinciding with elevated macroeconomic risks, as noted by the Tier-3 ranking for Italy in Euromoney Country Risk surveys:

17/9/16: The Mudslide Cometh for Your Ladder

One chart that really says it all when it comes to the fortunes of the Euro area economy:

And, courtesy of these monetary acrobatics, we now have private corporates issuing debt at negative yields, nominal yields...  http://blogs.wsj.com/moneybeat/2016/09/15/negative-yielding-corporate-debt-good-for-your-wealth/.

The train wreck of monetarist absurdity is now so far out on the wobbly bridge of economic systems devoid of productivity growth, consumer demand growth and capex demand that even the vultures have taken into the skies in anticipation of some juicy carrion. With $16 trillion (at the end of August) in sovereign debt yielding negative and with corporates now being paid to borrow, the idea of the savings-investment link - the fundamental basis of the economy - makes about as much sense today as voodoo does in medicine. Even WSJ noted as much: http://www.wsj.com/articles/the-5-000-year-government-debt-bubble-1472685194.

Which brings us to the simple point of action: don't buy bonds. Don't buy stocks. Hold defensive assets in stable proportions: gold, silver, land, fishing rights... anything other than the fundamentals-free paper.

As I recently quipped to an asset manager I used to work with:

"A mudslide off this mountain of debt will have to happen in order to correct the excesses built up in recent years. There is too much liquidity mass built into the markets devoid of investment demand, and too weak of an economy holding it. Everywhere. By fundamental metrics of value-added growth and organic demand expansion potential, every economy is simply sick. There is no productivity growth. There is no EPS growth, even with declining S down to waves of buy-outs. There is debt growth, with no capex & no EPS growth to underwrite that debt. There is a global banking system running totally on fumes pumped into it at an ever increasing rate by the Central Banks through direct monetary policies and by indirect means (regulatory shenanigans of ever-shifting capital and assets quality revisions). There is no trade growth. There is no market growth for trade. Neither supply side, nor demand side can hold much more, and countries, like the U.S., have run out of ability to find new lines of credit to inflate their economies. Students - kids! - are now so deep in debt before they even start working, they can't afford rents, let alone homes. Housing shortages & rents inflation are out of control. GenZ and GenY cannot afford renting and paying for groceries, and everyone is pretending that the ‘shared economy’ is a form of salvation when it really is a sign that people can’t pay for that second bedroom and need roommates to cover basic bills. Amidst all of that: 1% is riding high and dragging with it 10% that are public sector ‘heroes’ while bribing the 15% that are the elderly and don't give a damn about the future as long as they can afford their prescriptions. Take kids out of the equation, and the outright net recipients of subsidies and supports, and you have 25-30% of the total population who are carrying all the burden for the rest and are being crushed under debt, taxes and jobs markets that provide shit-for-wages careers. Happy times! Buy S&P. Buy penny stocks. Buy bonds. Buy sovereign debt. Buy risk-free Treasuries… Buy, Buy, Buy we hear from the sell-side. Because if you do not 'buy' you will miss the 'ladder'... Sounds familiar, folks? Right on... just as 2007 battle cry 'Buy Anglo shares' or 2005 call to 'Buy Romanian apartments' because, you know... who wants to miss 'The Ladder'?.."

14/9/16: Apple Tax on Keiser Report

My interview with Max Keiser for Keiser Report on Apple Tax fiasco: https://www.rt.com/shows/keiser-report/358886-episode-max-keiser-965/

14/9/16: Economics as Art

My conversation with Inspireland on Economics, philosophy of social studies and social inquiry:
https://inspirelandpodcast.com/2016/09/06/e47-constantin-gurdgiev-economics-is-art/

14/9/16: Expresso on ECB forward policy options

My comment on ECB's recent decisions (pre-meeting) in Portugal's Expresso
http://expresso.sapo.pt/economia/2016-09-07-Com-os-governos-ausentes-BCE-empurrado-para-fazer-mais

13/9/16: U.S. business investment slump: oil spoil?

Credit Suisse The Financialist recently asked a very important question: How low can U.S. business investment go? The question is really about the core drivers of the U.S. recovery post-GFC.

As The Financialist notes: “Over the last 50 years, there has usually been just one reason that businesses have slashed investment levels for prolonged periods of time—because the economy was down in the dumps.”

There is a handy chart to show this much.

“Not this time”, chimes The Financialist. In fact, “Private, nonresidential fixed investment fell 1.3 percent in real terms over the previous year in the second quarter of 2016, the third consecutive quarterly decline.” This the second time over the last 50 years that this has happened without there being an ongoing recession in the U.S.

Per Credit Suisse, the entire problem is down to oil-linked investment. And in part they are right. Latest figures reported by Bloomberg suggest that oil majors are set to slash USD1 trillion from global investment and spending on exploration and development. This is spread over 6 years: 2015-2020. So, on average, we are looking at roughly USD160 bn in capex and associated expenditure cuts globally, per annum. Roughly 2/3rds of this is down to cuts by the U.S. companies, and roughly 2/3rds of the balance is capex (as opposed to spending). Which brings potential cuts to investment by U.S. firms to around USD70 billion per annum at the upper envelope of estimates.

Incidentally, similar number of impact from oil price slump can be glimpsed from the fact that over 2010-2015, oil companies have issued USD1.2 trillion in debt, most of which is used for funding multi annual investment allocations.

Wait, that is hardly a massively significant number.

Worse, consider shaded areas marking recessions. Notice the ratio of trough to peak recoveries in investment in previous recessions. The average for pre-2007 episode is a 1:3 ratio (per one unit recovery, 3 units growth post-recovery). In the current episode it was (at the peak of the recovery) 1:0.6. Worse yet, notice that in all previous recoveries, save for dot.com bubble crisis and most recent Global Financial Crisis, recoveries ended up over-shooting pre-recession level of y/y growth in capex.

Another thing to worry about for 'oil's the devil' school of thought on corporate investment slowdown: slump in oil-related investment should be creating opportunities for investment elsewhere. One example: Norway, where property investments are offsetting fully decline in oil and gas related investment. When oil price drops, consumers and companies enjoy reallocation of resources and purchasing power generated from energy cost savings to other areas of demand and investment. Yet, few analysts can explain why contraction in oil price (and associated drop in oil-related investment) is not fuelling investment boom anywhere else in the economy.

To me, the reason is simple. Investing companies need three key factors to undertake capex:
1) Surplus demand compared to supply;
2) Technological capacity for investment; and
3) Policy and financial environment that is conducive to repatriation of returns from investment.

And guess what, they have none of these in the U.S.

Surplus demand creates pressure factor for investment, as firms face rapidly increasing demand with stable or slowly rising capacity to supply this demand. That is what happens in a normal recovery from a crisis. Unfortunately, we are not in a normal recovery. Consumer and corporate demand are being held down by slow growth in incomes, significant legacy debt burdens on household and corporate balance sheets, and demographics. Amplified sense of post-crisis vulnerability is also contributing to elevated levels of precautionary savings. So there is surplus supply capacity out there and not surplus demand. Which means that firms need less investment and more improvement in existent capital management / utilisation.

Technologically, we are not delivering a hell of a lot of new capacity for investment. Promising future technologies: AI-enabled robotics, 3-D printing, etc are still emerging and are yet to become a full mainstream. These are high risk technologies that are not exactly suited for taking over large scale capex budgets, yet.

Finally, fiscal, monetary and regulatory policies uncertainty is a huge headache across a range of sectors today. And we can add political uncertainty to that too. Take monetary uncertainty alone. We do not know 3-year to 5-year path for U.S. interest rates (policy rates, let alone market rates). Which means we have no decent visibility on the cost of capital forward. And we have a huge legacy debt load sitting across U.S. corporate balance sheets. So current debt levels have unknown forward costs, and future investment levels have unknown forward costs.

Just a few days ago I posted on the latest data involving U.S. corporate earnings (http://trueeconomics.blogspot.com/2016/09/7916-dont-tell-cheerleaders-us.html) - the headline says it all: the U.S. corporate environment is getting sicker and sicker by quarter.

Why would anyone invest in this environment? Even if oil is and energy are vastly cheaper than they were before and interest rates vastly lower...

12/9/16: Fiscal Policy in the Age of Debt

In recent years, there has been lots and lots of debates, discussions, arguments and research papers on the perennial topic of fiscal stimulus (aka Keynesian economics) on the recovery. The key concept in all these debates is that of a fiscal multipliers: by how much does an economy expand it the Government spending rises by EUR1 or a given % of GDP.

Surprisingly, little of the debate has focused on a simple set of environmental factors: fiscal stimulus takes place not in a vacuum of environmental conditions, but is coincident with: (a) economies in different stages of fiscal health (high / low deficits, high/low debt levels etc) and (b) economies in different stages of business cycle (expansion or contraction). One recent paper from the World Bank decided to correct for this glaring omission.

“Do Fiscal Multipliers Depend on Fiscal Positions?” by Raju Huidrom, M. Ayhan Kose, Jamus J. Lim and Franziska L. Ohnsorge (Policy Research Working Paper 7724, World Bank) looked at “the relationship between fiscal multipliers and fiscal positions of governments” based on a “large data-set of advanced and developing economies.” The authors deployed methodology that “permits tracing the endogenous relationship between fiscal multipliers and fiscal positions while maintaining enough degrees of freedom to draw sharp inferences.”

The authors report three key findings:

“First, the fiscal multipliers depend on fiscal positions: the multipliers tend to be larger when fiscal positions are strong (i.e. when government debt and deficits are low) than weak.” In other words, fiscal expansions work better in case where sovereigns are in better health.

“For instance, our estimates suggest that the long run multiplier can be as big as unity when the  fiscal position is strong but it can turn negative when the fiscal position is weak. A weak fiscal position can undermine fiscal multipliers even during recessions. Consistent with theoretical predictions, we provide empirical evidence suggesting that weak  fiscal positions are associated with smaller multipliers through both a Ricardian channel and an interest rate channel.”

By strong/weak fiscal position, the authors mean low/high sovereign debt to GDP ratio. And they show that fiscal expenditure uplift for higher debt ratio states results in economic waste (negative multipliers) in pro-cyclical spending cases (when fiscal expansion is undertaken at the times of growing economy). Which is important, because most of the ‘stimuli’ take place in such conditions and majority of the arguments in favour of fiscal spending increases happen on foot of rising economic growth (‘spend/invest while you have it’).

“Second, these effects are separate and distinct from the impact of the business cycle on
the fiscal multiplier.” Which means that debt/GDP ratio has an impact in terms of strengthening or weakening fiscal policy impact also regardless of the business cycle. Even if fiscal expansion is counter-cyclical (Keynesian in nature, or deployed at the time of a recession), fiscal multipliers (effectiveness of fiscal policy) are weaker whenever the debt/GDP ratio is higher. In a way, this is consistent with the issues arising in the literature examining effects of debt overhang on growth.

“Third, the state-dependent effects of the fiscal position on multipliers is attributable to two factors: an interest rate channel through which higher borrowing costs, due to investors’ increased perception of credit risks when stimulus is implemented from a weak initial fiscal position, crowd out private investment; and a Ricardian channel through which households reduce consumption in anticipation of future fiscal adjustments.”

What this means is that low interest rates (accommodative monetary policy) may be supporting positive effects of fiscal expansion, but at a cost of reducing private investment. In a sense, public investment, requiring lower interest rates, crowds out private investment. Now, no medals for guessing which environment we are witnessing today.

Some charts

First, median responses to increased Government spending

Once you control for debt/GDP position with stimulus taking place during recessions:

“Note: The graphs show the conditional fiscal multipliers during recessions for different levels of fiscal position at select horizons… Government debt as a percentage of GDP is the measure of fiscal position and the values shown on the x-axis correspond to the 5th to 95th percentiles from the sample. …Fiscal position is strong (weak) when government debt is low (high). Solid lines represent the median, and dotted bands are the 16-84 percent confidence bands.”

In the two charts above, notice that the range of public debt/GDP ratios for positive growth effect (multiplier > 1) of fiscal policy is effectively at or below 25%. At debt levels around 67%, fiscal expansion turns really costly (negative multipliers) in the long run. How many advanced economies have debt levels below 67%? How many below 25%? Care to count? Five  economies have debt levels below 25% (Estonia, Hong Kong, Macao, Luxembourg and San Marino). For 67% - nineteen out of 39 have debt levels above this threshold. Not exactly promising for fiscal expansions...

Overall, the paper is important in: (1) charting the relationship between fiscal policy effectiveness, and debt position of the sovereign; (2) linking coincident fiscal and monetary expansions to weaker private investment; and (3) showing that in the long run, fiscal expansion has serious costs in terms of growth and these costs are more pronounced for countries with higher debt levels. Now, about that idea that Greece, or the rest of PIGS, should run up public investment to combat growth crisis…

11/9/16: BRIC PMIs: Composite Activity - August

In the previous post I promised to update Composite PMI indicators for BRIC economies, so here it comes.

The good news is that Russia and India are posting Composite readings that are statistically significantly above 50.0 for the second month in a row. For Russia, this is the third consecutive month of Composite PMI readings statistically above 50.0 and for India - second.

The bad news is that Brazil acts as big drag on BRIC growth with severely depressed Composite PMI reading for 18th month running. Worse, Brazil's position has deteriorated in August compared to July.

Meanwhile, China posted virtually unchanged Composite PMI in August compared with July, with both readings being very close to signalling statistically significant expansion. Last time China posted statistically significant reading above 50 line was in August 2014.

Couple of charts to illustrate the trends:

As the chart above indicates, Russia remains a driver to the upside in terms of BRIC economies PMIs, with Brazil acting as a major drag and China as a driver toward lower growth.

Good news: across overall BRIC grouping, growth remains positive (albeit very shallow) and is ticking up (albeit with increased volatility). Bad news: since 1Q 2013, BRIC economies as a group are showing extremely low growth performance compared to their historical trends (red box in the chart below).

11/9/16: BRIC PMIs: Services & Manufacturing - August

With full 3Q 2016 update on PMIs coming up relatively soon, and having not done monthly updates on the time series for some time now, here is a quick summary of BRIC Manufacturing and Services PMIs through August 2016:

On Manufacturing side:

• Brazil remains firmly stuck under 50.0 and the talk about improvements in the economy is highly premature. The rate of contraction did slow down a bit in recent months, but getting worse more slowly is not equivalent to getting better. With 19 consecutive months of sub-50 readings, the manufacturing side of Brazil's economy remains deeply sick. Last time Brazil's manufacturing posted statistically significant growth was in March 2013. Ouch!
• Russia has been posting volatile manufacturing PMIs headlines for some time now. August reversion of PMI to 50.8 - statistically indistinguishable from 50.0 - offers no change to this pattern. That said, Russian Manufacturing appears to be stable, as opposed to contracting. Last 3mo average is at 50.6 - which, statistically, signals zero growth. This compares somewhat positively against 48.6 3mo average through May 2016. Overall, Russian manufacturing has not posted statistically significant growth reading - based on PMIs - since December 2014, with exception of one month (June 2016).
• China's Manufacturing PMI posted a non-contractionary reading of 50.0 (zero growth) in August, down from 50.6 in July. In statistical terms, Chines manufacturing posted contraction or zero growth readings for 25 consecutive months now.
• India continued to post significantly positive growth in manufacturing, based on PMIs. Over the last 8 months index reading stayed above 50.0 (statistically above 50.0 in 4 months out of 8). Current expansionary period in Indian manufacturing is now 8 months long and strengthening.

Chart below summarises trends in Manufacturing PMIs

The above shows that Manufacturing sectors are converging toward growth recovery in Russia and China, while India remains well-ahead of the rest of BRIC economies in terms of positive growth momentum. Brazil is on a clear downward trend and has decoupled from the other BRICs.

Services sectors:

As the above illustrates:

• Brazil services sectors posted yet another month of declining growth, with rate of decline accelerating in August compared to July. This marks 18th consecutive month of negative growth in the services sector in the country. As with manufacturing, country services have been performing extremely poorly since March 2013, when structural (long-term trend) slowdown in growth kicked in.
• Russia services sectors posted 7th consecutive month of above 50.0 readings, signalling relatively strong (albeit slower than in July) recovery. Over the last 6 months, Russia posted statistically significant growth in 5 months, which is rather solid sector recovery compared to the same period of 2015.
• Chinese services sectors never posted a reading below 50.0 in the entire history of the time series. However, in August, the series reading of 52.1 was stronger the July reading and marked the third time the series were statistically above 50.0 over the last 6 months. This suggests some firming up in the services sector growth in China - a welcome relief to the rather pessimistic outlook projected by the PMIs in previous months.
• India services PMI rose strongly to statistically significant reading of 54.7 in August, marking 14th straight month of above 50.0 readings (in level terms). August was the third month out of the last 6 months with statistically significant growth reading.

Just as with manufacturing, BRIC services sectors posted continuous improvements in trading conditions in India, China and Russia over the recent months. Brazil, however, remains significant drag on BRIC growth with no signs of convergence to the rest of the BRIC economies in sight.

Overall: both Manufacturing and Services PMIs suggest that BRIC economies as a group continue to act as a moderating factor on global growth trends. Although no longer dragging the global economy into growth recession, the block of largest emerging markets economies is not exactly propelling world growth to higher trend levels. However, more analysis on this later, with Composite indicators.