Sunday, 24 April 2016

Why Does the Bank of Canada Now Believe that Fiscal Stimulus Works?

Here are some quotes from a famous article in the November 1963 issue of the Canadian Journal of Economics and Political Science by Nobel Prize winning economist, Robert Mundell, then of McGill University. He is writing about the effect of a fiscal stimulus generated by an increase in government spending financed by borrowing in an economy with a floating exchange rate in a global environment of extremely high international capital mobility.
Increased [government] expenditure creates excess demand for goods and tends to raise income. But this would increase the demand for money, raise interest rates, attract a capital inflow and appreciate the exchange rate, which in turn would have a depressing effect on income. In fact, therefore, the negative effect on income of exchange rate appreciation has to offset exactly the positive multiplier effect on income of the original increase in government spending. Income cannot change unless the money supply or interest rates change. ... Fiscal policy thus completely loses its force as a domestic stabilizer when the exchange rate is allowed to fluctuate and the money supply is held constant. 
If my assumptions about capital mobility were valid in Canada, it would mean that expansive fiscal policy under flexible exchange rates [would be] of little help in increasing employment because of the ensuing inflow of capital which keeps the exchange rate high and induces a balance of trade deficit: we [would observe] a zero or very small multiplier. 
Of course the assumption of perfect capital mobility is not literally accurate... To the extent that Canada can maintain an interest rate equilibrium different from that of the United States, without strong capital flows, fiscal policy can be expected to play some role under flexible exchange rates... But if this possibility exists for us today, we can conjecture that it will exist to a lesser extent in the future.
Mundell's classic "Capital Mobility and Stabilization Policy Under  Fixed and Flexible Exchange Rates" clearly laid out that in a global economy with a high degree of capital mobility, fiscal stimulus in a small open economy operating with a flexible rate is likely to have little if any lasting impact on real GDP growth.

Fast forward 50 years and empirical studies by the US National Bureau of Economic Research (NBER) and by the Bank of Canada confirmed that fiscal multipliers for Canada are very low.

The NBER summarized its results as follows:
Based on a novel quarterly dataset of government expenditure in 44 countries [including Canada], we find that (i) the output effect of an increase in government consumption is larger in industrial than in developing countries, (ii) the fiscal multiplier is relatively large in economies operating under predetermined exchange rates but is zero in economies operating under flexible exchange rates; (iii) fiscal multipliers in open economies are smaller than in closed economies; (iv) fiscal multipliers in high-debt countries are negative. 
Specifically, the NBER found that for economies, like Canada, that are "open to trade or operating under flexible exchange rates, a fiscal expansion leads to no significant output gains. Further, fiscal stimulus may be counterproductive in highly indebted economies. Indeed, in countries with debt levels as low as 60% of GDP, government consumption shocks may have strong negative effects on output".

The Bank of Canada found that while Canada benefits greatly from fiscal stimulus conducted by the United States and other G20 trading partners, "Canada’s high level of openness sharply curtails the effectiveness of a domestic [fiscal] stimulus." Specifically, the 2010 study by BoC researchers found that, for Canada, fiscal multipliers for government spending were very low: ranging from virtually zero for increases in general transfer payments, to 0.40% for government investment expenditures, to 0.80% for government spending on goods, to .98% for spending on government services.

The findings of the NBER and the BoC, corroborate Mundell's theoretical insight and support his view that as global capital markets became ever more efficient, fiscal policy has become ever less powerful in small open economies with flexible exchange rates.

So it is somewhat surprising to me that in recent days, the Governor and the Senior Deputy Governor of the Bank of Canada have gone out of their way to tell members of the press that they believe that Canada's latest fiscal stimulus will have a substantial positive effect on real GDP.

Governor Stephen Poloz told the Canadian Press, "we're fortunate in Canada that we have that fiscal capability right now to shift our policy mix ... as the government has done... This is exactly the setting where fiscal policy is its most effective and its also where monetary policy is its least effective." 

Bloomberg News reported Poloz as saying that the size of the fiscal multiplier depends on the “situation you start in.” When the economy is in equilibrium, government spending triggers higher interest rates and higher currency, crowding out private spending. In equilibrium, fiscal policy “has no effect except it changes the mix of what is going on in the economy. So this gives rise to sort of cavalier statements from some that would say, 'well that won’t have any effect.'" Poloz went on to say that in low growth settings, “fiscal policy begins to add demand in the economy (and) basically nothing else happens except that demand goes up and what happens then is that you get the maximum effect of fiscal policy because there are no offsets such as upward creep in interest rates or movement in the exchange rate.”

BoC Senior Deputy Governor Carolyn Wilkins was singing from the same hymn book in an interview with Macleans magazine, where she opined that fiscal policy is more effective than monetary policy in boosting productivity and growth when interest rates are low. In the interview, she said "If fiscal policy can do some of the heavy lifting, that’s a positive thing. Fiscal policy at low interest rates is also just more effective. In a world where growth is going to be structurally slower because of demographic changes, monetary policy can’t fix that. If we want sustainable growth, we need to boost productivity, not only in Canada [but in] the global economy. That’s the only place growth can come from."

Mr. Poloz seems to be basing his enthusiasm for fiscal policy on the degree of slack in the Canadian economy. Indeed, some studies do indicate that government spending multipliers for OECD countries, on average, are near zero in periods of expansion or low unemployment but are higher in recessions or in periods of very high unemployment. But Canada is not in recession and unemployment is close to the threshold level of 7%, only above which fiscal multipliers might be somewhat higher. And Canada is not an average OECD country; it is a smaller, much more open economy than the US, the Eurozone or Japan and it is much more sensitive to movements in its exchange rate.

Mr. Poloz and Ms. Wilkins seem to be arguing that one of Robert Mundell's Nobel prize winning theoretical insights does not apply to the Canadian economy, the very economy where Mundell pointed out that his theory was most likely to be true. They seem to be arguing that the 2010 empirical results of the NBER and of the BoC's own researchers, which showed zero or very low fiscal multipliers are not to be believed.

I am not convinced. On the contrary, I believe that Mundell's theory is more valid today than it ever was. The economy has some slack as Poloz suggests, but not so much, as by the BoC's estimate it was operating somewhere between 98.3% and 99.4% of full capacity in the first quarter of 2016. One can concede that fiscal stimulus may lead to a short run pickup in growth. However, the market's perception of improved short-term growth even before the budget measures were implemented and the market's anticipation of a further fillip to growth from fiscal stimulus has already raised interest rate expectations compared to what they were when the BoC was still perceived to be in easing mode prior to January 20, when the BoC signalled that it would stand pat. The higher interest rate expectations for Canada, combined with more accommodative monetary stances by the US Fed, the ECB and the BoJ, have already caused the Canadian dollar to appreciate by 15% against the US dollar from its January low. The stronger Canadian dollar has already caused the BoC to downgrade its export growth forecast. As the fiscal stimulus feeds into the economy in 2016 and 2017, growth of private sector production for export markets will be suppressed by the stronger Canadian dollar and will be supplanted by increased public spending on infrastructure and on low multiplier transfer payments. With Canada's G20 trading partners not moving to large-scale fiscal stimulus, the result will be what the BoC researchers predicted in 2010: Canada’s high level of openness and flexible exchange rate will "sharply curtail the effectiveness of domestic fiscal stimulus."

It also seems to suggest that the stance of the BoC's monetary policy relative to our major trading partners is still a very powerful policy instrument.

The eventual outcome of the fiscal stimulus being applied in a non-recessionary economy without similar actions in Canada's trading partners is unlikely to be sustained stronger economic growth. Instead, the cumulative impact on the level of real GDP is likely to be negligible. However, the composition of real GDP and the level of government debt will be affected. 

Real GDP will be shifted away from private sector output of goods and services toward public sector construction output and subsidized green energy projects. With private sector construction of pipelines and LNG terminals stalled by government regulations, and with coal-fired power plants being shut down, productivity seems more likely to fall than to rise, as Ms. Wilkins argues.

Public debt to GDP is already rising and will continue to rise. According to the National Balance Sheet Accounts, total government debt rose to 76.2% of GDP at the end of 2015 from 72.8% a year earlier. Recent federal and provincial budgets suggest that the increase in total government debt to GDP will accelerate over the next few years. This could put Canada quickly into the situation described by the NBER researchers in which fiscal multipliers in high debt countries may have strong negative effects on real GDP.

Finally, even if one believed that fiscal stimulus would work to boost real output and employment on a sustained basis [which I don't], one can only wonder why an independent Bank of Canada did not voice this opinion a year ago when the economy was actually contracting, but then praised stimulus after the newly elected government tabled a budget promising much higher spending and deficits.

Sunday, 3 April 2016

Global ETF Portfolios for Canadian Investors: 1Q16 Review and Outlook

After being a terrible strategy in 2015, Canadian investors did well to stay at home in 1Q16.

In 2015, the major forces driving markets were the plunge in crude oil prices and the monetary policy divergences between the US Fed, which announced its intention raise its policy rate, and other major central banks, including the Bank of Canada, which were expected to ease monetary policy in response to weak growth and below target inflation. These forces shifted in 1Q16.

The price of WTI crude oil, which began 1Q16 at US$37 per barrel, touched a 12-year low of just over US$26/bbl on February 11, but then rallied back to finish the quarter at US$38/bbl. The rally in crude oil was triggered by two factors. First, the market perceived, and the Fed confirmed, that it would move more gradually to raise its policy rate. This led to profit-taking on long-USD positions and sent the US dollar index (DXY) down by 5% from its late-January high to its March 31 low. The weaker USD, combined with an agreement between Russia and Saudi Arabia to cap oil production at current levels, helped boost crude oil and other commodity prices. 

On the monetary policy front, in 2015, as the Fed promised that it would raise the fed funds rate, market participants viewed the Bank of Canada as one of the group of central banks that were expected to ease policy further. However, while the central banks of Japan, the Eurozone, China, Sweden, Norway and New Zealand did ease further in 1Q16, the BoC failed to validate the market's expectation of further easing, deciding to defer any further action until after the federal budget provided details of the new Liberal government's promised fiscal stimulus. The BoC decision to stand pat, combined with the Fed backing off on tightening, sparked a 12% rally in the Canadian dollar from the January 19 low of 68.6 US cents to the March 31 high of 77 cents.

This 12% rally was hard on unhedged Canadian portfolios with substantial global exposures. For example, as the Fed signalled a more gradual pace of tightening, the S&P500 ETF (SPY) reversed its early January losses: it gained 10% in USD terms from January 15 to March 31, but lost 1.5% in CAD terms. Similarly, the US 10-20 year Treasury bond ETF (TLH) gained 3% in USD terms over the same period, but lost 7.8% in CAD terms. 

As a result of the sharp C$ appreciation, all of the unhedged Global ETF portfolios that I track in this blog posted losses for 1Q16. Meanwhile, a stay-at-home 60/40 investor who invested 60% of their funds in a Canadian stock ETF (XIU), 30% in a Canadian bond ETF (XBB), and 10% in a Canadian real return bond ETF (XRB) had a total return (including reinvested dividend and interest payments) of 2.6% in Canadian dollars. This was a partial reversal of the 3.0% loss posted by the all-Canada portfolio in 2015.

Since we began monitoring at the beginning of 2012, we have found that the unhedged Global ETF portfolios have vastly outperformed the stay-at-home portfolio. However, as you will see in this post, that behaviour reversed in 1Q16.

Global Market ETFs: Performance for 1Q16

In 1Q16, with the USD depreciating 6% against the CAD (after appreciating 19% in 2015), the best performing global ETFs in CAD terms were an unlikely trio: gold, local currency emerging market bonds and Canadian equities. The worst ETF returns were in the Japanese and Eurozone equities and commodities. The chart below shows 1Q16 returns, including reinvested dividends, for the ETFs tracked in this blog, in both USD terms and CAD terms. 

Global ETF returns varied widely across the different asset classes in 1Q16. In USD terms, 14 of the 19 ETFs we track posted positive returns, while 5 ETFs posted losses for the quarter. However, just 9 of the 19 ETFs posted gains in CAD terms, while 10 posted losses. 

The best gain was in the Gold ETF (GLD) which returned 9.3% in CAD terms. The Emerging Market Local Currency bond ETF (EMLC) returned 3.8%, while the Canadian equity ETF (XIU) returned 3.4% in CAD terms. Other ETFs with positive returns for the quarter included Canadian Long Bonds (XLB) 2.8%; Canadian real return bonds (XRB) 1.6%; non-US sovereign government bonds (BWX) 1.6%; non-US inflation-linked bonds (WIP) 1.4%; Canadian corporate bonds (XCB) 1.3%; and the emerging market equity ETF (EEM) 0.1% in CAD terms. 

The worst performers, were the Japanese equity ETF (EWJ) and the Eurozone equity ETF (FEZ), which returned -11.4% and -9.2%, respectively in CAD terms. Other ETFs posting sizeable negative returns in CAD terms were the commodity ETF (GSG) -8.8%; US small cap stocks (IWM) -7.3%; the S&P500 ETF (SPY) -4.7%; the US high yield bond ETF (HYG) -3.7%; US inflation-linked bonds (TIP) -1.7%; and US Investment Grade Bonds (LQD) -1.4%.

Global ETF Portfolio Performance for 1Q16

In 1Q16, the Global ETF portfolios tracked in this blog all posted negative returns in CAD terms when USD currency exposure was left unhedged. However, if the USD currency exposure was hedged, all of these portfolios would have realized positive returns. 

A stay-at-home, Canada only 60% equity/40% Bond Portfolio returned 2.6%, as mentioned at the top of this post. Among the global ETF portfolios that we track, the Global 60% Equity/40% Bond ETF Portfolio (including both Canadian and global equity and bond ETFs) returned -2.6% in CAD terms when USD exposure was left unhedged, but +0.9% if the USD exposure was hedged. A less volatile portfolio for cautious investors, the Global 45/25/30, comprised of 45% global equities, 25% government and corporate bonds and 30% cash, lost 1.6% if unhedged, but gained 1.7% if USD hedged.

Risk balanced portfolios underperformed in 1Q16 if unhedged, but outperformed if hedged (the reverse of their performance in 2015). A Global Levered Risk Balanced (RB) Portfolio, which uses leverage to balance the expected risk contribution from the Global Market ETFs, lost 1.3% in CAD terms if USD-unhedged, but had the biggest gain of +5.8% if USD-hedged, benefitting from strong levered bond returns in USD terms. An Unlevered Global Risk Balanced (RB) Portfolio, which has less exposure to government bonds, inflation-linked bonds and commodities but more exposure to corporate credit, lost1.3% if USD-unhedged, but gained 3.5% if USD-hedged.

Looking Ahead

In my view, the key market events of 1Q16 that influenced global ETF portfolio returns in CAD terms were: the Fed's shift to a less aggressive tightening stance; the BoC's decision to defer further easing; and the agreement between the Russians and the Saudis to cap oil production at current high levels.

As we look ahead, it is fair to say that prospective returns on these portfolios will depend on how views on monetary policies and crude oil prices evolve in the coming months. 

Recent economic data suggest that 1Q16 real GDP growth will turn out to be stronger than expected in Canada and weaker than expected in the US. It should be noted however that US economic data is more timely than Canadian data. Early indications for 1Q16, based on the Atlanta Fed's GDP Now forecast, pointed to close to 3% growth for US real GDP, but that estimate has now been cut to 0.7%. Similarly, recent upward revisions to 1Q16 growth forecasts to 3% for Canada have been heavily influenced by a strong gain in January GDP. It would not be surprising if these expectations are tempered somewhat, just as US forecasts have been cut as more current data has been released. Nevertheless, the combination of the bounce in Canada's GDP combined with substantial increases in government spending contained in the federal budget, will likely keep the Bank of Canada on hold for some time. Meanwhile, Fed Chair Janet Yellen's message has been clear that the Fed will proceed cautiously with tightening, so the shift in monetary policy expectations that occurred in 1Q16 is likely to persist.

A sharp decline in worldwide oil and gas exploration and development spending suggests that crude oil production growth will slow, perhaps bringing global oil demand and supply into better balance as 2016 unfolds. While crude oil prices have come off their late March highs, they are unlikely to dip back below US$30/bbl any time soon.

This leaves markets in a similar position to where they were at the beginning of 2016. Both equity and bond market valuations remain stretched. Global growth remains sluggish. Deflationary forces remain strong. China's credit growth remains worrisome. Geo-political risks remain elevated.

As I wrote three months ago, "in a continuing uncertain environment, characterized by significant global divergences in growth and central bank policies, and depressed oil and other commodity prices, remaining well diversified across asset classes, with substantial exposure to USD-denominated assets and with an ample cash position continues to be a prudent strategy". While this strategy generated modest losses in 1Q16 as the Canadian dollar surged, it still seems prudent to me.