Pages

Wednesday, May 23, 2012

Capital Inflows Across Asia

A new IMF working paper looks at capital inflows and what can be done about them (abstract):

Surging Capital Flows to Emerging Asia: Facts, Impacts, and Responses
Balakrishnan, Ravi and Sylwia Nowak; Sanjaya Panth & Wu Yiqun

Summary: Net capital flows to emerging Asia rebounded at a record pace following the global financial crisis, raising concerns about overheating and financial stability. This paper documents the size and composition of the most recent surge to Asian emerging markets from a historical perspective and compares developments in the broader economy, asset prices, and corporate variables across the different episodes of strong inflows. We find little evidence of a significant build-up of imbalances and resource misallocation during the most recent surge. We also review country experiences in managing the risks associated with inflows and argue that Asian countries have used regulatory measures during past surges, although there is not strong evidence of their efficacy without supporting monetary and fiscal policies.

There’s some interesting differences in experiences between different categories of Asian emerging markets.

Why should we be concerned? Because massive inflows run the risk of “sudden stop” episodes where capital inflows either stop, or worse reverse course entirely in a very short period. That could potentially put extreme pressure on exchange rates, and financial market and banking liquidity.

In short, it’s not good for macroeconomic stability or business confidence. The study actually identifies that approximately 60% of capital inflow episodes for emerging Asia in their sample range (1987-2011) ended in a sudden stop – which is an uncomfortably high percentage.

Moreover, we’re also looking at inflows that have become shorter in duration over time i.e. greater volatility, which isn’t good either as it limits the policy response menu (i.e. you can’t put in place medium-to-long term policies). The flip side to that is it also means that there’s less likelihood of high capital inflows changing long term investment and borrowing behaviour, or causing asset price bubbles.

The NIEs (Hong Kong, Korea, Singapore and Taiwan) appear to be most highly exposed to capital flow volatility in recent time – as of 2011, their volatility is double that of ASEAN, and 5 times that of China.

Specific to Malaysia, there are two points of vulnerability identified, both relating to asset prices: house prices  which continue to remain elevated relative to incomes, and foreign holdings of government bonds. While the former is somewhat ameliorated by low price appreciation prior to the Great Recession, the latter is real enough (ratio of foreign holdings to total government debt):

01_rentas

MGS is a pretty big and liquid market, and has been preferred over equities by foreign investors. There’s no doubt that foreign investment in Malaysian bonds have helped keep government borrowing costs perhaps artificially lower than it should be. Of the other major Asian economies, only Indonesia is more vulnerable this way, with the ratio is exceeding 30%.

In any case, the paper is a nice, quick and digestible survey of capital inflows and outflows in Asia over the last 25 years.

Technical Notes:

Balakrishnan, Ravi and Sylwia Nowak; Sanjaya Panth & Wu Yiqun , "Surging Capital Flows to Emerging Asia: Facts, Impacts, and Responses", IMF Working Paper No. 12/130, May 2012

6 comments:

  1. I read Dr. Ravi's earlier paper (2011) on the same subject. Feels a bit the same.

    Link to pdf: http://www.imf.org/external/pubs/ft/sdn/2011/sdn1110.pdf

    My 2 cents though, Indonesia seems a bit strange. (Relatively) strict capital controls with somewhat low domestic liquidity yet there's still so much foreign funds.

    Smells very, very fishy. If the Indonesian growth story ever slows down, high interest rates would probably cripple them following a massive capital flight - controls notwithstanding.

    ReplyDelete
  2. Jason, they've bought lots of "insurance". Their reserves are rapidly catching up to Malaysian and Singaporean levels. Plus, the exchange rate is fairly flexible, which should act as a good shock absorber. Our mistake 15 years ago was being overly wedded to a soft peg against the USD.

    ReplyDelete
  3. It's good news that they do have their reserves, but their Current Account looks comparatively weak (and it looks like it might dip into deficit territory soon).

    I'd argue that the Rupiah has been less flexible than the ringgit since 2005 and its sudden depreciation may in fact accelerate interest rate pressures rather than increasing its FX buffers. But I guess that's a separate discussion.

    Further, domestic monetary policy transmission channel is seen to be very laggy. Data has it that interest rate announcements barely changes domestic bank rates.

    That said, I'm not too sure whether their level of reserves are able to mitigate the effects of the ensuing upward interest rate pressures associated with the possible outflows.

    *at the end of this post I realize I might be going overboard with Indonesia. Will redirect the discussion back to something more closer-to-home*

    ReplyDelete
  4. :)

    BTW, "Data has it that interest rate announcements barely changes domestic bank rates."

    Check the data for Malaysia. It barely changes here either.

    ReplyDelete
  5. Erm.. we might have different data sets for our local banks then.

    BNM reports that the weighted average lending rate have a 2-month lag (after any OPR adjustment) prior to the crisis. Post-crisis, OPR changes and lending rate changes seemed to have moved in tandem.

    Still, that's not a very clear indication whether the transmission mechanism is effective locally.

    Great questions during the DOS briefing btw.

    ReplyDelete
  6. Thanks.

    We're looking at the same data, probably through different perspectives. The interest rate pass-through is really incomplete:

    http://econsmalaysia.blogspot.com/2011/07/nontransmission-of-monetary-policy.html

    ReplyDelete