Thursday, October 17, 2013

How To Spin With Statistics

Here’s a model article for all budding analysts and pundits out there (excerpt):

Malaise Is Ahead For Malaysia's Bubble Economy

I recently wrote about how Indonesia’s economy has devolved into a classic credit and asset bubble-driven growth story, and its neighbor Malaysia is on the same path along with most other Southeast Asian economies, which are part of the overall emerging markets bubble that I have been warning about in the last couple of years.

The emerging markets bubble began in 2009 after China pursued an aggressive credit-driven infrastructure-based growth strategy to bolster their economy during the global financial crisis. China’s economy quickly rebounded as construction activity flourished, which drove a global raw materials boom that greatly benefited commodities exporting countries such as Australia and emerging markets. Emerging markets’ improving fortunes began to attract the attention of global investors who were seeking to diversify away from Western nations that were at the epicenter of the financial crisis.

Rock-bottom interest rates in the U.S., Europe, and Japan, combined with the Federal Reserve’s multi-trillion dollar quantitative easing programs encouraged a $4 trillion torrent of speculative “hot money” to flow into emerging market investments over the past four years. A global carry trade arose in which investors borrowed at low interest rates from the U.S. and Japan, invested the funds in high-yielding emerging market assets, and pocketed the interest rate differential or “spread.” Soaring demand for EM assets led to a bond bubble and ultra-low borrowing costs, which resulted in government-driven infrastructure booms, alarmingly fast credit growth, and property bubbles in numerous developing nations…

His main thesis is that QE and loose monetary policy in the West has flooded emerging markets, and fostered asset and property bubbles through cheap credit. And he builds an ostensibly persuasive case with lots of data and charts.

Here’s the trick:

Ultra-low interest rates have caused Malaysia’s private sector loans to increase by over 80 percent since 2008…

Malaysia’s M3 money supply, a broad measure of total money and credit in the economy, shows a similar worrisome trend:

Malaysia’s household credit bubble is helping to fuel a consumer spending boom:

Malaysian car registrations are up by 50 percent since 2008:

Accounting for nearly half of all household debt, soaring mortgage loan growth is a primary reason why Malaysia’s household debt is increasing at such a rapid rate.

Malaysia’s rapidly deteriorating current account surplus due to weaker exports is another worrisome development.

Every one of the charts accompanying these comments are in levels, not in annual % change. You don’t see trends, all you see are big numbers getting bigger. The comparisons are all in total change from a given year, which naturally inflates the numbers you present.

Here’s what happens when you actually check out the trends:

  1. Credit growth has dropped from 12% in 2011 to under 10% this year, and household borrowing growth has similarly dropped from 13% to about 11% currently. Fast growth? Credit growth in both Singapore and Indonesia is at least double that;
  2. Ditto M3 growth, which peaked at 14.7% in Feb 2012, and is now below 8%;
  3. Private consumption growth has fluctuated between 8% and 6%, below the 8%-10% growth rates seen before the 2008-2009 crisis;
  4. Car registration rates dropped from 13.9% in 2010 to 1.6% in 2011, before rising slightly to 5.6% in 2012.
  5. Actually, personal loans from non-banks is the primary reason for higher household debt. And the higher household debt to GDP ratio is also partly due to slowing NGDP growth.
  6. The current account surplus is deteriorating largely due to higher imports of capital goods. Exports have been flat, not really declining.

In other words, the facts and data are creatively presented in such a way as to support the case he’s trying to make.

Dude, the bubble’s already deflating. You’ve missed it.

I also had to laugh at this one:

Since 2010, Malaysia’s public debt-to-GDP ratio has been hovering at all time highs of over 50 percent thanks to large fiscal deficits that were incurred when an aggressive stimulus package was launched to bolster the country’s economy during the Global Financial Crisis.

Malaysia’s all time high is actually 103% in 1986. The “aggressive stimulus package” was neither aggressive, nor particularly stimulating.

So if you’ve got an argument that you want backed up by “facts”, just remember there are lots of ways to package data up even if it says something quite different.


  1. Dear Hisham

    Actually, personal loans from non-banks is the primary reason for higher household debt. And the higher household debt to GDP ratio is also partly due to slowing NGDP growth. "

    wow. this is surprising.. what kind of debt is this? when ppl buy iPhone by installments? appliances at court mammoth? Those kinda debt? pls do shar ethe data. I am intrigued by this. From Rafizi's analysis.. a constant point that soaring house prices and over inflated car prices are the cause of household debt increase.

    1. What METAL said, though Muamalat doesn't count (it's an Islamic bank, not DFI).

      Details on what's going on with non-banks are here.

  2. Likely DFIs like MBSB Bank Rakyat BSN Bank Muamalat

  3. how do we contain soaring house prices? is the current house prices at a fair level or is it artificial fueled by speculation? if we can contain house prices, surely our household debt will be in check..

    1. It's difficult to contain house price increases, as the market isn't liquid and prices are determined on the margin i.e. even one sale at an inflated price sets a new price benchmark.

      Loan restrictions only have limited impact (from historical experience in other countries), so the next best option is increase the supply. To be fair though, banks have been steadily rationing mortgage credit since 2006 - approval ratios have dropped from over 60% then to under 50% now.

    2. Hi Anon,
      How did you come to that conclusion? That's quite a leap.

      On containment of soaring house prices, 3 ways to approach it:
      1) Rid of speculative demand. Mechanisms include stiffer RPGTs (in principle, I support capital gains taxes of any form in general) and Stamp Duties. These mechanisms are surgical in reducing demand from a targeted segment (speculators). But I doubt they are the main drivers of home price inflation.

      2) Reduce money supply. Credit expansion is fuel for inflation. Mechanisms include increasing bank capital requirements and increasing interest rates. (Blunt instruments with lots of side-effects). BNM may be tempted to intervene with floor price controls on mortgages but that would be unthinkable. Tightening lending policies also sort of works but this one is really a mix of preventing new demand and hence new credit expansion.

      3) Increase housing supply. There are about 4.7m houses in Malaysia today discounting kampung houses in a nation of 29m. According to Rehda, the rate of new formation of households is 140k a year, while new houses is about 100k a year. Of course REHDA's claims must always be taken w/ a pinch of salt but I believe there is some truth to fundamental demand being there.

    3. It takes too long to increase house supply in Malaysia to be able to damage control in time. Changing rules to speed up will only encourage rouge developers to produce unsafe houses or apartments (see what happens in India and Bangladesh where housing collapsing are common events). Only option is through sales rules, i.e. no foreigners allowed to buy second hand property , or properties in certain locations etc. More houses in Malay Reserve only land will help bumi buyers as these properties cannot be sold to non bumis hence priced lower..Provided no hanky panky allowed.

    4. On the point of no foreigners allowed to buy second hand property, I doubt that will have much of an impact. Currently property purchases by foreigners represent 2% of transaction, and even then it's for properties above RM500k

    5. Metal, on point 3, new completions for housing in Malaysia

      2003: 261k
      2004: 209k
      2005: 191k
      2006: 199k
      2007: 195k
      2008: 146k
      2009: 132k
      2010: 110k
      2011: 77k
      2012: 119k

      Growth in residential stock has fallen from over 8% in 2003, to 2.6% in 2012.

      All data from NAPIC

    6. Could it be that condos are no longer known as condos, but service apartments under commercial land, thus a commercial stock rather than a residential stock?

      Soho, sofo, sovo... Whatever creative funny names they come up with, it is for residential purpose on commercial land title.

    7. @anon 5.14

      That's a really interesting idea and there may be some truth in that. I'll need to do some digging on this.

    8. Destroy Greedy Developers!October 27, 2013 at 3:33 AM

      house prices can be limited by limiting loans limits. Just like when the Govt raised car loans to 55k for govt servants, proton simply raised the price to 55k. If there are no loans for 400k houses then develpers will find ways to build prices to 250,000. Why? simply there are more buyers at the lower level. What about land price, cement and steel? Its the same effect. We were living in attap houses for centuries. The current wasteful cement and mortar are just choices not necessity.
      The people must be aware of the greed process for profiteering at all sectors in the supply chain.

  4. The real injustice is really income inequality exacerbated by loose lending policies and unfair taxation.

    At present, increases in capital gains far outruns increases in income from labor.

    So those who can afford their first home will be able to afford many many more. Those who can't will never be able to afford one.

    1. Metal,

      "The real injustice is really income inequality exacerbated by loose lending policies and unfair taxation"

      I think it is more likely the influx of foreign workers that suppress wages as well as globalization (china made goods) that keep the income level of worker low or almost no real gain over the last decade.

      Zuo De

  5. Hishamh,

    I was re-reading your article about the stimulus that was neither aggressive nor stimulating in particular to the multiplier effect...

    The first BR1M i thought was quite stimulating in terms of multiplier effect (resulting in strong Q2, 2012 GDP growth) but the second time round the effect was much muted, is this due to the (higher) base effect from the previous year? Or was all this a myth.

    Zuo De

    1. Zuo De,

      Base effect. If I give you RM500 for the very first time, there should be a measurable effect on growth. If I give RM500 to you a second time, will that turn up in growth or a continuation of the previous level of spending?

    2. Yup understood, how about the multiplier, none infact maybe negative?

      Zuo De

    3. Zuo De,

      The general consensus in the most recent literature is that fiscal multipliers for consumption are positive but small.

  6. Basically Jesse Colombo ideas is he trying to warn any economic bubble's that may happen in future (Although I know he maybe negatively speaking).

    But he knows the basic thing of money creation (Money is created by commercial banks through debt) thus economic crisis will happen once in a decade.

    So we do not want our country being trap again (Post 2008 global financial crisis Malaysia have to inject RM67.0 billion into the economy and added it to our debt level).

    I am only a Master candidate in Economics at local university in Malaysia. I still learn untill now.

    And I know there is thousand valid reasons/explanations by just looking at certain data.

    But when there goes to fiat money, creating money out of thin air, economic crisis all the points will merge into only one valid reason.

    1. @anon 8.59

      The problem is not the message, the problem is the way it is delivered.

      There is no real attempt at analysis, no attempt at a balanced view, no discussion of where his viewpoint might be wrong. In academia, this is called data mining - take a viewpoint and only present the data that represents that viewpoint.

      It's dishonest.


      "...Malaysia have to inject RM67.0 billion into the economy and added it to our debt level..."

      By my calculations, the net injection was only about RM2b.

      "But when there goes to fiat money, creating money out of thin air, economic crisis all the points will merge into only one valid reason."

      This ignores the multiplicity of crises under hard money systems. The point is, whichever monetary system is in use has to accommodate the fact that economies are made up of many people, each with rational and irrational behaviours and beliefs.

      Unless you believe such economies are self-correcting and self-regulating, any monetary system will be subject to periodic crises.

  7. Well, it seems that Malaysia is in pretty good company. Besides the Malaysian bubble, real or imagined, Mr Colombo is also warning about "growing bubbles in Canada, Australia, Nordic countries, China, emerging markets, Web 2.0 startups, U.S. higher education, and more." I am rather shocked that he missed out the kid with the soap water and blow thingie on my street corner, who still seems to be producing the largest and greatest number of bubbles of all!

    1. @anon 12.04


      Yes, the problem with one trick ponies is that they only that one trick...

  8. Mind sharing what are the sources for your data in this article?

    1. @anon 4.54

      I use primary sources. Try the link at the top right hand corner of this page (second from top).

      BTW, let me know if you find any dead links.

      The only exception is car registration numbers, which can be found on the JPJ website