The Global Week Ahead: Is A US Housing Recovery Sustainable?

Next Week's Risk Dashboard

• Why US housing may be on a sustained rebound
• Canadian CPI could make or break BoC hike bets
• Canadian GDP to post mild growth
• BoC surveys to inform inflation expectations
• US PCE to follow CPI higher
• Eurozone CPI unlikely to sway the ECB’s next decision
• ECB’s Sintra conference
• Australian CPI may impact the RBA’s next decision
• China’s PMIs to inform downside risks
• BanRep likely to hold, too early to talk cuts
• Riksbank to hike by a minimum 25bps
• Other global macro readings

Chart of the Week

Several housing markets around the world are beginning to come back to life. House prices are back on the upswing in places such as Australia and Canada. Is the US housing market in a sustained recovery? This is among the more important matters to the US economic outlook and the potential global market consequences given the sheer enormity of the US economy and its powerful central bank. A solid case can be made for answering in the affirmative, but not necessarily with uniform drivers, and even while the housing data is a bit puzzling on balance. Here’s my take on how to decipher the developments by walking through the readings before getting to a concluding narrative.

If housing is in recovery, then mortgage lenders are not seeing much by way of the fruits thus far. Mortgage purchase applications are largely treading water (chart 1), but there could be more positive signs emerging in other indicators and on the road ahead.

So far, existing home sales are doing likewise, although this masks some regional gains. Pending home sales are a leading indicator of existing home sales by 30–90 days but are also trending sideways albeit lagging a month (chart 2). This lag is only because the US records existing home sales when the keys exchange (versus, say, Canada that records them when the contracts are initially signed).

And yet, house prices may be bottoming and back on an upswing as evidenced by the US repeat sales home price gauge and housing starts have very recently soared (chart 3) which is where the narrative begins to twist.

One indication of why this is happening is that prospective homebuyers' foot traffic through model homes has been sharply rising for several months. This is a leading indicator of new home sales and the reason why the salespeople working in model homes jots down your visits.

New home sales have also been trending upward and are highly correlated with foot traffic as evidence of the appetite for home ownership by tire (brick?) kickers (chart 4).

So how do we square the circle in terms of what's going on overall?

One possibility is that with the caution that we need more data to assess, it's feasible that what's holding back existing home sales figures is a shortage of inventory as indicated by months' supply riding under 3 months (chart 5). In order to find product, buyers are funnelling into the new build segment. If so, then this is a bigger positive for growth since there is far more value-added for GDP purposes attached to building homes and to swapping resales that only offer limited value-added on the related support services.

Another possibility is that folks already in the market don't want to move and reset their lucrative 30-year mortgage rates higher and so this is impairing supply in the resale market. After all, the effective paid mortgage interest burden on outstanding mortgage debt is very, very low right now by historical standards (chart 6) and American borrowers hold a one-way option to refinance when mortgage rates fall without the obligation to do so when rates are rising unless they move. For anyone in search of an existing home, this inventory shortfall is causing buyers to pay up. For new entrants into the housing market this isn't an issue. Clearly some can afford the 30-yr rate and they have little choice but to go into new homes.

Hence, why homebuilder stocks are doing rather well despite 500bps of Fed hikes to date (chart 7). In essence, their share prices are performing well because the funnelling of incremental volume demand into newly built housing is dampening the rate sensitivity of housing.

A key support is that the cost of financing a home purchase in the US is much cheaper than would appear to be the case by only looking at the 30-year nominal fixed mortgage rate. First, we have to deflate the cost of borrowing and one way of doing that is by using expected house price changes (chart 8). This approach trades-off the cost of servicing mortgage debt against the expected capital appreciation. Another point is that Americans can deduct mortgage interest depending upon how they stack up in relation to the changes introduced by the 2017 Tax Cuts and Jobs Act. The TCJA changes no longer require itemization of deductions which also helps to reduce the effective burden of mortgage debt. As mortgage interest and deductions rise, new home buyers can punt some of the effects of higher inflation back to the US Treasury and in essence someone else ultimately pays for it and quite possibly themselves in the longer run!

So why is all of this not yet showing up in mortgage applications and mortgage credit figures? Applications may rise soon in lagging fashion. Perhaps some of the buying and building is by cash investors. As for credit growth landing on the books, if the momentum is shifting in a sustained way into the new build segment then there's a lag during construction until it shows up in mortgages being granted.

This is all very tentative, but with robust US household finances marked by around record-low debt payments to incomes, locked-in ultra-cheap mortgages, a 22-year low in the household debt-to-income ratio, high cash balances, very strong job markets and pent-up demand, the conditions may be ripe for expanding the housing stock alongside price gains even absent material overall population growth. If the US ever improves in terms of immigration policies, then the outlook would only become brighter yet.

CENTRAL BANKS—ENJOY THE CANAPÉS

The scope for major central banks surprising markets will slip into a quieter period until next month when many of them issue their next batch of decisions before taking August off. A couple of central bank stragglers will bring us fresh decisions next week after the onslaught of decisions that markets have digested of late. The ECB will host its annual photo-op for the biggest global central bankers. Light information from the Bank of Canada may also be impactful.

ECB’s Sintra—Fêting the Rock Stars

The ECB’s annual Forum on Central Banking will be held once again in beautiful Sintra, Portugal starting on Monday through Wednesday. These central bankers sure know how to pick their paid vacation destinations. Next up is the Fed’s Jackson Hole conference in late August set against the gorgeous backdrop of the Grand Teton mountain range (agenda here). There is value to these gatherings as the sharing of ideas and discussions can have the participants coming away with fresh insights. The optics around the gatherings are likely to be more closely scrutinized by the rest of the world that’s dealing with their collectively botched management of inflation, the lagging pressures on borrowing costs and the highly regressive consequences.

The agenda is here. The ECB’s Lagarde, Fed’s Powell, BoE’s Bailey, BoJ’s Ueda, Norges Bank’s Bache, and a host of other global central bank officials will participate in a variety of sessions. It’s doubtful that we’ll hear something meaningfully new given that the four biggest names—in alphabetical order, Bailey, Lagarde, Powell and Ueda—will share 90 minutes after having spilled the beans with their recent policy decisions.

BanRep—Not So Cool Inflation

Banco de la Republica de Colombia delivers its latest policy decision and guidance on Friday. After surprising the market at its April meeting by raising its lending rate by 25bps, a majority of analysts are expecting the central bank of Colombia to hold its lending rate at 13.25%.

Does that mean the risk of surprise is now gone? Perhaps. Officials have guided they expect inflation to begin easing over 2023H2. It’s not showing such signs to date. True, year-over-year rates of inflation are ebbing, but as elsewhere, this is significantly driven by year-ago base effects. Month-over-month gains remain abnormally strong and may extend a hawkish bias. Chart 9 demonstrates this point by showing that the officially provided seasonally unadjusted gains in core inflation over May and the prior month of April were among the strongest on record when combined with like months over history. This is a better way of looking at price pressures at the margin. Until this begins sustainably turning weaker, expectations for policy easing would appear to be fanciful.

Riksbank—Fading April’s Words

Sweden’s Riksbank delivers its latest policy decision on Thursday. Forward guidance that was provided at the April 26^th meeting explicitly stated that the central bank would hike either at this coming meeting or the next one in September. Developments since then may have put a more exigent spin on the timing, magnitude and forward guidance.

For one, the krona has depreciated by about 3½% which risks fanning imported inflation. For another, inflation recently surprised higher with core inflation remaining sticky at around 7% y/y. Chart 10 shows the preferred way of looking at inflationary pressures not in year-over-year terms that can be driven by base effects, but in month-over-month terms using source data that is seasonally unadjusted and hence compared to like months in history. April and May this year both registered well above typical inflation for like months in the past, indicating that underlying inflationary pressures remain robust.

Furthermore, the external environment has changed. Several foreign central banks have increased their hawkishness including the Federal Reserve but also central banks closer to home including 50bps hikes by the Bank of England and Norges Bank alongside continued tightening guidance from the ECB.

Because of these considerations, markets have moved toward sitting on the fence between pricing a 25bps and 50bps hike from the Riksbank this time and therefore treating prior guidance as stale by now.

BoC—Survey Says!?

Not this again. Please. Uncle!

The Bank of Canada normally saves its surveys of businesses and consumers for release on the Monday of the week before the week that contains policy decisions accompanied by a Monetary Policy Report and hence a full forecast update. Not this time. The surveys arrive on Friday because the following Monday is Canada Day and hence a statutory holiday.

The surveys tend to be on the stale side, but the BoC has made efforts to supplement the results with some fresher perspectives. We’ll hear about businesses’ plans for hiring, wage increases, cap-ex investment, expected sales growth and, most importantly of all for BoC watchers, what they expect for inflation 1-year, 2-years and 5-years ahead. The prior edition that was released in early April showed that businesses don’t believe the BoC will hit its 2% inflation target over the entire horizon (chart 11).

We’ll also hear about what consumers expect for inflation rates with the same recent disbelief in the 2% target (chart 12), plus expectations for house price appreciation and wage gains among other factors.

Thus, the heart of the matter here is the evidence on the ongoing degree to which inflation expectations remain unmoored. Canada has very limited gauges to help assess this issue, not least of which since measures derived from financial markets via real return bonds are unreliable given high liquidity distortions and more recently the decision to stop issuing them which makes little sense to me.

Also note that Deputy Governor Sharon Kozicki speaks on Tuesday. I think it will offer very low risk to BoC watchers. She is a discussant on a panel session titled “Assessing the costs of inflation” that is focused upon an academic paper titled “Inflation and misallocation in New Keynesian models” at the ECB’s Sintra, Portugal conference on central banking. There will be no media and no published remarks. But hey, Portugal can be a beautiful place to visit.

CANADIAN INFLATION—THIS COULD MAKE OR BREAK A JULY HIKE

Canada updates CPI inflation for the month of May on Tuesday. The release kickstarts a focus upon the fundamentals including the next day’s GDP report and Friday’s inflation expectations within the BoC’s surveys as noted above.

This one could matter a lot to BoC pricing with markets pricing a strong chance at a 25bps hike in July and about a half-point hike before year-end. That’s because this will be the last inflation report that we’ll get before the BoC’s next decision on July 12^th and so a surprise in either direction could influence whether additional tightening unfolds, whether it proceeds in a straight line, or whether tightening involves skips along the way as it has in some prior cycles especially at the fine-tuning stages. One example of this was in early 2015 but in the opposite direction to today’s bias; the BoC controversially cut in January, skipped, then cut again. June CPI doesn’t land until about a week after the BoC’s July decision after which we don’t get another BoC decision until September. Other releases along the way will also matter of course, such as the next round of data on jobs, wages and activity readings.

I’ve pencilled in a rise of 0.5% m/m for headline CPI in seasonally unadjusted (NSA) terms that would translate into a mild seasonally adjusted gain of 0.1–0.2% m/m. Base effects will drag the year-over-year CPI inflation rate down from 4.4% toward 3% before other adjustments. The usual caution against being fooled by base effects continues to apply, versus focusing upon price pressures at the margin (ie: m/m SA core gauges).

May is usually a mild seasonal up-month for prices. Gasoline prices were flat in May over June in seasonally unadjusted terms and should offer no contribution this time but may return as a modest upside in June. Gasoline prices will probably knock about 0.7 ppts off of year-over-year total CPI.

Traditional core CPI (ex-food and energy) is expected to climb by 0.5% m/m NSA but in this case that translates into a 0.4% m/m SA gain.

What will matter more to the BoC will be evidence on the continued breadth of gains (chart 13) and the month-over-month trend pressures on the trimmed mean and weighted median core inflation gauges. The fact that these measures accelerated in April was among the drivers of the BoC’s decision to hike by 25bps in June rather than waiting (chart 14). It’s difficult to attempt to provide credible estimates of these gauges for the month of May since we need the full distribution of all prices given how sensitive the measures are to exactly what is included. The weighted median CPI measure captures the change in prices for the good or service at the 50^th percentile of the weighted distribution of changes and can therefore be super-sensitive to that midpoint of the distribution. The trimmed mean gauge weeds out the top and bottom 20% of prices and readjusts the weights for the remainder to come up with the overall gauge. Some shops provide estimates, but it’s largely futile in a bottom-up short-term sense given how unobservable many of the components are in advance.

This release will incorporate updates to basket weights to reflect 2022 spending levels and the composition of spending versus 2021. How they affect shelter costs may be a risk. The adjustments affect CPI from May forward with no adjustments to history. There were also a few adjustments to what is included with the curious additions of snow removal equipment and baby products since I guess they didn’t previously exist…and the rather lagging removal of DVD players.

I don’t expect a significant impact on CPI from the basket weight changes. Chart 15 shows the adjusted weights by category. The weight on shelter was dropped by about 1½ percentage points and was cancelled out by increased weights on other components. The net effect on year-over-year rates had these weights been used instead of the 2021 weights until April is shown in chart 16 and is inconsequential. The impact on month-over-month shelter costs could be a risk because of the 1.5 ppt decline in the weight absent retroactive weight adjustments and given that new house prices are recently stabilizing (chart 17). Canadian CPI captures house prices using the house-only component of the new house price index and the weight adjustments lowered the weight on this component by of shelter costs 1.1 ppts.

Finally, this piece from the BoC also factors into the discussion on inflation. With plenty of caveats that are noted by the authors, they conclude that businesses have been passing on cost increases roughly one-for-one without raising their markups. That leans against the charge that ‘greedflation’ and changed corporate pricing behaviour is driving overall inflation. That doesn’t mean there is no ‘greedflation’ whatsoever but leans toward downplaying it as an overall driver of broad price pressures at least until the folks levying such assertions come up with better evidence to prove their claims in an evidence-based field of inquiry.

GLOBAL INFLATION—A HIGH BAR IS SET FOR THESE TO MATTER

Canada won’t be the only one updating inflation readings this week, but its update will matter more to the next round of central bank decisions than the other updates may matter to their respective central banks.

US INFLATION—Too Early to Matter

US core PCE inflation during May arrives on Friday. I’ve estimated a headline rise of 0.1% m/m SA and a core PCE increase ex-food and energy of 0.4% m/m. These estimates are broadly in line with the already known CPI figures and the Cleveland Fed’s ‘nowcast’ PCE estimates.

PCE could matter relatively less to the Fed’s July 26^th decision than CPI to the BoC’s July 12^th decision.

One reason for this is that we already know what happened to core CPI in May and it was up by 0.4%. Core PCE tends to closely follow core CPI but deviations are possible given different methodologies and weighted schemes. For instance, core PCE’s much lower weight on shelter costs than in CPI could mean that the expected disinflation in that component over 2023H2 and beyond won’t have as large of an effect on core PCE.

Another reason is that the FOMC will take down another CPI report on July 12^th ahead of the 26^th decision which should give a decent impression of what their preferred PCE reading will look like two days after the July decision. There would have to be material improvement compared to the established stickiness to knock them off course from further tightening (chart 18).

EUROZONE INFLATION—The ECB Already Sounds Committed

Friday’s Eurozone CPI figures for June might also matter less to the ECB in a relative central bank sense than to the BoC. True, this is the final CPI print before that July 27^th decision given that July CPI arrives only on the 31^st, but the ECB has largely already provided guidance it intends to hike again on July 27^th and there is likely a high bar set against knocking them off course. It would take more than one such surprise to do so given the trend of stronger core CPI gains than normal like months in the past that saw some progress in May. Furthermore, accelerating wage gains are driving concerns around the outlook for trend inflation going forward (chart 19).

Seasonally unadjusted gasoline prices were little changed in June over May and therefore may be a relative drag on the month’s overall price changes. After controlling for this effect, consensus still lands in a fairly uncertain range of estimates from 0.2% m/m to 0.5% m/m CPI increases and with year-over-year core CPI holding at 5.3%. What will matter critically is whether June core CPI repeats the pattern of stronger than usual seasonally unadjusted gains compared to like months in history.

Advance clues will be offered when Italy releases CPI on Wednesday, followed by Germany and Spain on Thursday such that there may be relatively little intrigue left over by the time the Eurozone tally arrives along with France’s estimates.

AUSTRALIA—The Last One before the Next RBA Decision

Australia started to publish monthly estimates of CPI only last September after previously sticking to quarterly updates. The next monthly estimate will be released for the month of May on Tuesday. Like Canada’s update, this one could matter to the RBA’s next decision on July 4^th given it’s the last one before then and with markets pricing only a modest chance at a 25bps hike.

There is a wide range of guesswork within consensus but most estimates land within a range from 5.8% y/y to 6.3% from 6.8% the prior month. Key will be incremental core price pressures in month-over-month terms and breadth.

We don’t get the trimmed mean measure with the monthly figures after the Australian Bureau of Statistics suspended the monthly estimates in December. Instead, it offered to keep the focus upon quarterly trimmed mean while leaving it to markets to evaluate the substitute measure for CPI excluding volatile items and holiday travel, as well as the compositional breadth of the price changes. CPI ex-volatile items and holiday travel serves as a higher frequency close proxy for quarterly trimmed mean CPI (chart 20).

GLOBAL MACRO—IS CHINA’S ECONOMY STILL SOFTENING?

The global release calendar will include several nuggets focused upon Chinese growth, Canada’s economy, several US readings and light European developments.

Asia-Pacific markets will be primarily focused upon the latest batch of state purchasing managers’ indices for June on Thursday that may further test downside risks in China’s economy (chart 21). Also due out are Australian retail sales (Wednesday), and Japan’s monthly sweep of indicators across the household and industrial sectors (Thursday). 

Canada will update GDP for April with details and for May sans details on Friday. Statcan previously offered ‘flash’ guidance that GDP was up 0.2% m/m in April. Since then, the evolution of the data has my simple regression equation of GDP against known macro readings anticipating a tick higher than that. May GDP is likely to be soft given we already know that hours worked fell 0.4% m/m and that matters because GDP is the product of hours times labour productivity. Activity readings that would play into productivity estimates are tracking fairly softly with, for example, softer construction alongside firmer retail volumes. On net, GDP may be tracking around 1 ½% q/q at a seasonally adjusted and annualized pace using the monthly figures. After strong Q1 growth, Q2 growth may be riding in tandem with the slowly expanding supply side of the economy which means no progress as yet toward reducing excess demand in the Canadian economy. Only once this process begins can we begin lagging out disinflationary effects toward the BoC’s 2% goal at least insofar as its output gap framework is concerned. There are nevertheless plenty of other drivers of inflation than gaps.

US releases will be relatively light over the coming week, apart from the potential for Friday’s PCE inflation figures to be impactful. That same day is likely to bring out mild gains in income and spending during May. Durable goods orders (Tuesday) are expected to be soft but orders ex-transportation may be firmer. Another batch of repeat sales house price figures for April will attempt to build a positive trend (Tuesday) before the next day’s possible gain in new home sales during May and then Thursday’s pending resale home figures. The Conference Board’s measure of consumer confidence (Tuesday) could follow the UofM sentiment gauge higher. Also watch for a possible mild upward revision to Q1 GDP from the prior 1.3% annualized growth estimate (Thursday) and lesser gauges summarized in the accompanying tables.

European markets will primarily focus upon Eurozone CPI with just French consumer spending during May (Friday) and a few German releases to consider including IFO business confidence (Monday), and then both retail sales and the unemployment rate (Friday).