The March US housing report was the first month since 2008 showing starts at over one million units on a seasonally adjusted annual rate (SAAR). It confirmed that the recovery in US housing markets was continuing in early 2013.
In 2010, most months recorded annual starts below a 0.6 million rate; in the first four months of 2013, starts have averaged 0.935 million – about 60% above 2010 and 29% higher than the same period in 2012.
However, the starting point for this recovery was unusually low. Prior to the Great Recession, US housing analysts would have assumed a cyclical low of 0.9 to 1.0 million.
Arguably, US housing starts in early 2013 have simply returned to ‘normal’ cyclical lows and have a long way to go before reaching ‘normal’ average levels.
Given the lower levels of immigration for several years, most estimates of underlying annual demand for new homes in the US now fall in the range of 1.5 to 1.7 million units. Consequently, the early 2013 construction pace still has a long way to go to reach average annual underlying demand.
US housing market conditions have continued to improve in the spring of 2013; foreclosures are tumbling; home prices are rising; home sales are strong; and inventories of unsold new and existing homes are very low relative to actual, and potential, demand.
In economically buoyant areas, the lack of homes for sale is a major constraint on increased sales, encouraging new home construction. However, tight lending requirements; lack of available building land; and tightness in labour supply are all limiting the pace of recovery in new homes.
However, with continued growth in the overall US economy and rising employment, as well as rising incomes and continued low mortgage rates, the housing recovery is likely to be sustained over at least the next two years.
This outlook is reflected in Table 1 and Figure 1. For 2013, the CFPA forecast looks for US housing starts to total 1.06 million (36% above 2012), to climb to 1.45 million in 2014 and to reach the 1.60 million level in 2015.
In addition, US mobile home production will climb from around 55,000 units in 2012 to over 100,000 by 2015. Total US new housing production could therefore jump from 0.84 million units in 2012 to 1.72 million units in 2015 – a 100%-plus increase over three years.
Canada Partially offsetting the positive story in US housing is the weakening in Canadian housing since 2012. Housing starts in 2012 reached 215,000 units, only slightly below 2005 and 12% above the 2010-2011 average (Table 1). Fears of a housing bubble in Canada have led to tightening of mortgage lending, reflected in a substantial drop in Canadian housing starts.
In January-April, 2013, Canadian housing starts averaged just 175,000 a year – down 19% on January-April 2012. CFPA has revised its housing forecast lower for the next few years to reflect changes in Canadian housing policies, as well as the less-certain economic environment for an economy ever-more dependent on commodity, rather than manufacturing, exports.
For 2013, Canadian housing starts of 0.178 million will be 17% below 2012, but are expected to edge higher through 2015, to 0.192 million. The combination of lower Canadian starts and rising US housing production will drop the Canadian share of North American new home construction to 11% by 2014-15, down from 22% in 2012.
Total North American conventional housing starts (excluding mobile homes) will climb 24%, from 0.996 million in 2012 to 1.238 million in 2013, as US strength far outweighs Canadian weakness. Further substantial growth is expected over the following two years to 1.79 million in 2015, 80% above 2012. This growth in the key North American end-use market for wood products is already reflected in the improvement in both volumes and prices for both softwood lumber and panels.
A cautionary note has to be sounded, given the volatility in monthly housing starts data and the changes in the house-type mix. The US April housing report revealed a very large drop (17%) in seasonally-adjusted starts as the very volatile multi-family sector witnessed a 38% plunge from March levels. However, that plunge followed a 26% jump in March!
Monthly changes in the important singlefamily sector were smaller and, year-on-year, US single-family starts in April were up 21%. It will continue to be the strength (or weakness) in single-family starts that will be the chief market driver for lumber and panel producers. It is encouraging that the drop in the single-family share of new residential construction has slowed and may be finding a bottom as rising rents force people to consider owning rather than renting their homes.
The single-family share of US new home construction in 2005 was 83%, but by 2012 had dropped to 68%. Further declines to 66% in 2013 and 63% by 2015 are anticipated in the CFPA forecast. Reflecting the most recent data, this projection is somewhat more bullish compared to last year’s forecast, thus further supporting growth in wood products demand.
As we have noted before, the issue for the North American wood products market is not the absolute level of US housing starts relative to historical peaks. Given the attrition in panel and lumber capacity during the Great Recession, it is the year-to-year change in housing starts (and thus panel and lumber demand) in relation to the lower operating capacity that is fundamental to market analysis. Relatively small year-on-year increases in total wood products demand in the second half of 2012/early 2013 led to big jumps in North American lumber, OSB and softwood plywood prices (and smaller increases for MDF/HDF).
Market direction was reversed this spring. A relatively cold winter/spring held back the seasonal recovery in construction, meaning growing product inventories in the distribution pipeline. Consequently, after soaring in late 2012 and early 2013, many commodity wood product prices tumbled in April/May.
For example, between the end of February and mid-May, fob-mill softwood lumber prices fell 16%; OSB tumbled 25%; while southern pine plywood fell just 2% (Table 2). However, the plywood pricing behaviour noted in Table 2 is a little deceptive; plywood prices peaked later in the spring than OSB, and at mid-May were 14% off their mid-April highs.
In contrast with the typical volatility of commodity lumber and structural panel prices, the movements in MDF/HDF and particleboard have been a lot more subdued and displayed steadier growth over the past year. MDF/HDF has exhibited the greatest strength, but in recent months particleboard prices, particularly in the southeast, have strengthened and as of mid-May were up 7% over end-February.
Markets: Looking to the future The spring 2013 setback in lumber and panel markets is likely to be temporary, provided housing starts continue to rise. Continued volatility in commodity lumber and structural panel prices can almost be guaranteed as the ramp-up in production at times lags that in demand and at other times outpaces the increase in consumption.
Nominal dollar price peaks over the next two to three years will almost certainly be higher than the levels recorded in early 2013 and will likely be higher than the cyclical peaks of 2003-05. (In inflation-adjusted terms, prices in 2013-15 may not reach the earlier heights). And as excess capacity in particleboard and MDF/HDF becomes more utilised, further price rises are to be expected – especially for particleboard, where profitability has long been thin or non-existent. While the next three years will experience significant pricing volatility around an upward trend, the story for consumption, production and imports will be steadier growth.
Propelling the recovery in North American production will be housing- and repair and renovation-led increases in panel and lumber consumption. Given the strong links between US new home construction and OSB demand, the strongest element of the increase in total North American panel demand will be OSB.
Nevertheless, plywood will receive some spill-over benefits from housing’s healthier performance, as will MDF/HDF and particleboard. However, increased furniture production to furnish new homes, additions and upgrades will see the largest increase in demand for particleboard, as well as a major contribution to higher MDF/HDF consumption.
From just 33.8 million m3 in 2011, total North American panel consumption is forecast to climb to 39.6 million m3 in 2013 and to reach 47.8 million m3 in 2015 – 41% above 2011 (Figure 2). Despite this large increase, 2015 projected consumption will still be 17% off the 2005 pace, largely because of weaknesses in plywood and particleboard.
What do these increases in consumption mean for domestic North American production and imports? As through much of recent history, North America (US and Canada) will continue to account for the bulk of domestic panel consumption (as well as exporting some product to Mexico, Asia, Europe and the Middle East). In recent years, imports have represented just 5% of the combined North American supply of OSB, softwood plywood, MDF/HDF and particleboard (1.70 million m3). At the 2005 peak, imports were 8% of total supply (4.65 million m3) as the US imported large volumes of plywood as well as MDF.
Similar, albeit less dramatic, trends were reported for domestic North American production over the same period: From 53.1 million m3 in 2005, total panel production slumped 39%, to 32.1 million m3 in 2011.
Total production of North American panels is expected to climb to 37.5 million m3 in 2013 and to reach 44.2 million m3 in 2015 (31% above 2013 – Table 3). Meanwhile, total imports are projected to jump 106% from 1.72 million m3 in 2012 to 3.55 million m3 in 2015.
The respective growth rates for production and imports will reflect: (a) the capacity utilisation rates for each type of panel in North America; (b) the rate of capacity growth as mothballed operations are re-opened and older mills are modernised and de-bottlenecked; (c) exchange rates; and (d) how high domestic North American prices climb.
Looking at each panel type separately, there are clearly different trends and outcomes that can be expected for each type of panel. These trends are illustrated in Figure 3.
The capacity of OSB will certainly expand in the coming years. Announcements of reopening of mothballed mills make it clear that operating capacity at the end of 2015 will be substantially higher than at the end of 2012.
In addition, mills which were previously operating sub-optimally have either returned, or are returning, to full-time operation. What is unclear is the pace of the ramp-up of these reopened mills and whether unexpected difficulties in re-opening long-closed mills will delay expected increases in output.
Consequently, CFPA expects US imports of OSB from Europe and South America to reappear on the radar screen as US producers struggle to keep up with increased demand.
Plywood capacity will likely increase, but primarily through upgrading and debottlenecking operating mills, rather than from re-opening closed ones. Reluctance by domestic producers to re-open closed plywood mills will reflect several factors.
First, the limited amount of benefit to plywood from the rebound in housing starts; plywood sheathing only meets approximately 15% of structural panel demand in new construction.
Consequently, growth in softwood plywood demand will be modest over the next few years. Second, higher US prices will attract supply from Brazil and Chile, especially with the completion of the second line at CMPC’s operation by mid-2013 and Arauco’s rebuild of its Nueva Aldea mill by year-end. Re-opening closed plywood capacity in North America could be like shooting yourself in the foot! The sane, cautious approach would be to increase production modestly at existing operations through investments in efficiency.
Particleboard capacity in North America continues to well-exceed consumption and the imbalance is particularly acute in the US West. Arguably there should be more permanent closures by this industry, but tightening markets in the south and east may help to keep all current mills in the west operating as they ship larger volumes out of their region and into mid-west and Texas markets.
Imports remain a relatively small threat to particleboard producers unless domestic price levels rise substantially faster than costs over the next two years. While improved pricing and profitability are to be expected, it is unlikely this improvement will be substantial enough to attract significant import volumes. MDF/HDF capacity utilisation has tightened over the past year. While several mills have yet to ramp-up to seven day operation, overall the capacity slack is not great and very little closed capacity is available to be re-opened.
Consequently, it has to be expected that attention will be paid to building additional new lines at existing facilities or perhaps to constructing a greenfield mill, provided attractive fibre supplies and energy costs are available. Growth in domestic consumption, along with the opportunity to displace imported products with domestic supplies, represents a major opportunity for investment in the MDF/HDF industry in coming years. The longer domestic investment response lags, the greater the increase in imports to meet domestic consumption of panels, mouldings and laminate flooring. MDF/HDF will thus remain the largest contributor to total panel imports, at least through mid-decade. In summary, the outlook for North American panel markets continues to brighten.
Too rapid capacity- and production-growth could derail the recovery in pricing/profitability (as could a flatter growth curve in US home construction). But on balance the expectation is for continued improvement in volumes, pricing and profitability over the next two years, albeit with continued marked volatility in commodity structural panel markets.