The Strange And Mysterious Workings Of Singapore’s Monetary Policy

Singapore is a pretty unique economy, what with being a very open island trading nation, and with its political and social history.

Its approach to monetary policy is just as unique. Unlike the vast majority of central banks, Singapore’s Monetary Authority (MAS) uses the exchange rate as its primary monetary policy instrument. While this in itself is not too radical, unlike exchange rate regimes in the past the application of this policy is not through targeting a level of the exchange rate, but the slope and breadth of its appreciation.

This makes economic sense, as inflation is an appreciation in the general price level, not the price level itself. If the goal of monetary policy is stable prices (and/or economic growth), then a policy of exchange rate appreciation to regulate an increase in prices is appropriate.

This is especially true when you consider that because Singapore produces little of its own consumables (limited arable land for food production is one example; fresh water is another constraint, desalination notwithstanding), inflation is largely due to imports and not domestic production.

The result is a monetary policy regime that is quite different from virtually everyone else’s.

One consequence of this is that, given Singapore’s high long term growth rate, the effective result is that MAS is constantly having to accumulate international reserves to contain the appreciation of the exchange rate, as the latter would be a de facto tightening of monetary policy. This has led to charges of currency manipulation, which is a bit ridiculous. Of course there’s manipulation, that’s the whole point.

Now the reason why I’m writing this post is another aspect of Singapore’s policy approach that is just as radically different, which is its approach to liquidity management.

Before I get into that however, I need to make a small digression. Over to Prof Balding (excerpt):

What if Temasek Actually Earned 17%?

…According to Temasek at the end of the fiscal year in March 2012 (Temasek has not yet released their March 31, 2013 annual report), they reported $198 billion SGD under management. If we assume that Temasek has exactly $198 billion SGD and not one penny more or less and that they earned exactly 17% and 17.01% or 16.99% this implies they began with $508 million SGD (increasing the range to 17.44% and 16.5% does not fundamentally alter this analysis).

This actually reconciles rather closely with other available information rather closely in two ways…

…Doing this however only creates a new and enormous problem.

From 1974 to 2012, Singapore enjoyed total operational public surpluses from $307 billion SGD and incurred new borrowing of $381 billion SGD. This means that between 1974 and 2012, Singapore had a total of $688 billion SGD in free cash flow for investment purposes.

Here is where we encounter the problem if we assume that Temasek earned 17% since 1974. In its public balance sheet ending March 31, 2012, the Singapore government lists total assets of $765 billion SGD. If we subtract out the $198 billion SGD managed by Temasek we are left with $567 billion SGD. If Temasek earned 17% annually from 1974 to 2012 that means that the government was receiving $688 billion SGD to somehow end up with $567 billion. In other words, the non-Temasek public Singapore investors managed to lose $121 billion SGD or about .5% annually.

Let us take this analysis one step further and assume that the public surpluses and borrowing earned a modest 5% after expenses and costs (the 5% number is for illustrative purposes only but less than 7% claimed by GIC). If the yearly surpluses and borrowing were invested every year and earned 5% annually, this would yield a total of $1.3 trillion SGD. The difference between the actual Singaporean non-Temasek assets and what Singapore should have if surpluses and borrowing since 1974 earned a conservative 5% is a staggering $750 billion SGD. In other words, Singapore is about $750 billion short of what it should have if Temasek earned 17% and the remain money earned 5%!

Though the evidence fails to support the claim that Temasek earned 17% annually since its inception in 1974, even if this is true it only creates bigger problems. If Temasek did legitimately earn 17% annualized what happened to the rest of this so called investment juggernaut?

I’ve touched on this particular subject before (here and here), but I want to delve a little deeper into this issue in this post.

Most central banks, using an inflation targeting regime and an interest rate policy instrument, manage liquidity through open market operations. This is sometimes done through the buying and selling of government securities (effectively the way quantitative easing is being done in the West), or more normally, through the issuance or buy back of central bank securities.

In Malaysia, Bank Negara issues BNM Bills through auctions, with maturities between 1-month to 12-months. Issuance of securities takes money out of the banking system, and those moneys are kept at the central bank, sans any return. The interest cost of these securities will need to be paid out to the bond holders, without any commensurate return to compensate (central banks earn their revenue in other ways).

These central bank securities are never taken as part of the government’s overall debt burden. I repeat, never. They are securities fully backed by cash and not invested. That’s the whole point of the exercise – to reduce the amount of money available in the financial system. Investing the monies raised through such means thoroughly defeats the intended purpose.

I would term this “monetary” debt, not “fiscal” debt. In fact, I find it hard to think of this as debt at all.

Here’s where Singapore again differs substantially from the norm; they mix both fiscal debt and monetary debt under the same class of securities. Instead of issuing debt from MAS, the practice has been to issue short term T-bills from the government. While the buying and selling of existing government securities have often been used to manage liquidity, the difference here is that Singapore appears to be specifically issuing short term government debt for this purpose. This is slowly changing, as MAS was granted the authority to issue its own debt in 2010.

Nevertheless, only about a fifth of outstanding short term official debt is currently MAS bills, and the majority is still in the form of short term Treasury securities, i.e. the official debt of the government, to the tune of SGD162 billion as of March 2013. That’s approximately 40% of total outstanding government debt.

Historically, the ratio has been closer to 50%, and has only come down over the last few years as MAS bills began to displace T-bills as the liquidity management tool of choice. More to the point, since money raised through T-Bill issuance is sitting in the government’s account at the central bank, it earns precisely zero.

Where Prof Balding is getting his numbers wrong is assuming that all the debt issued by the Singapore government is being invested or managed by Temasek or GIC. The truth is, half the debt issued over the years just sits at the central bank costing the government the prevailing short term interest rate but yielding precisely nothing. That would change the calculation of estimated returns substantially. In fact, without going through the numbers myself (I can’t report any workings as my data source is work-related), I suspect almost all the discrepancy disappears.

Unfortunately, MAS and the Singapore government have been unwilling, or unable, to explain this particular intersection of Singapore’s monetary policy and government debt at all. In fact, I don’t see many other economists picking up on this nuance either, with most taking the official line that Singapore’s official debt issuance is for “the development of the capital markets.”

That may be partially true, but juxtaposing the structure of Singapore government debt with conventional central bank liquidity operations, and you come up with a much more plausible and rational explanation.