LARGE FORMAT RETAIL Linking the Housing Market to LFR

Size: px

Start display at page:

Download "LARGE FORMAT RETAIL Linking the Housing Market to LFR"

Holly Hood
5 years ago
Views:

COLLIERS RADAR LARGE FORMAT RETAIL Linking the Market to Daniel Lees Director Research daniel.lees@colliers.

are implemented. Our analysis demonst that a relationship between the housing market and sector exists, specifically, an inverse relationship between mortgage rate movement and rental growth.

1 COLLIERS RADAR LARGE FORMAT RETAIL Linking the Market to Daniel Lees Director Research Summary The Large Format Retail () sector has come under an increasing level of scrutiny, as regulatory measures, designed to address medium term risks associated with the domestic housing market are implemented. Our analysis demonst that a relationship between the housing market and sector exists, specifically, an inverse relationship between mortgage rate movement and rental growth. However, while the establishment of such a relationship paints a subdued outlook for the sector on face value, our analysis also demonst that the upward movement in mortgage required to meaningfully reduce rents is material, and unlikely in the medium term. In this regard, the outlook for the sector in terms of rental growth is far less bearish that some recent media headlines suggest. How do mortgages impact rents? Although it is difficult to create a direct correlation between the movement in mortgage and the performance of rents, it is still possible to map out a series of events that would reasonably imply a relationship between the two. Essentially our thesis is that mortgage influence the availability of capital flowing into the residential market, which is reflected within the measure of housing (transactions involving existing housing stock, and lending associated with the construction and purchase of dwellings). As housing picks up, so too do the sale of household goods such as furniture, electronics, whitegoods, hardware items, building materials and garden equipment. In turn, increased household goods sales lead to increased earnings for tenants in this sector, enabling them to expand and pay higher rents. Since 2013, the below relationship has broadly held true: However, a possible risk facing the sector, is the reversal of this relationship, as policies are introduced to alleviate price inflation within the domestic housing market. In this case the opposite chain of events could play out:

2 Connecting the dots between housing and While the direct relationship between mortgage and rents is not statistically significant, the relationship between other variables is significant, especially when looking a higher frequency data points over longer periods. 1. The dependency of housing on mortgage To examine this relationship, we looked at housing. Specifically, this refers to the purchase of existing dwellings, rather than the construction of dwellings of the purchase of new dwellings. The number of transactions within the pool of existing housing stock is far higher (86% of total) than those being constructed (8% of total), or those that are new (6% of total). As part of housing, we also look at lending finance, specifically finance used for the purchase and construction of new dwellings, which is a much larger pool of funds, relative to finance used only for alterations and additions. For mortgage, we specifically used the standard variable lending and 3-year fixed for owner occupier mortgages. These have the longest time series (back to 1959 and 1990 respectively) vs investors (only since 2015). Our charting reveals the negative relationship between these factors. Note the mortgage rate time series are inverse, which is supported by our regression analysis implying that almost 73% of the movement in housing is explained by the movement in mortgage. National housing transactions (#) vs mortgage (%) 60,000 55,000 50,000 45,000 40,000 35,000 30,000 25,000 20,000 15,000 10, Source: ABS / RBA National housing transactions (LHS) Own occ var % (inv) (RHS) Own occ 3Y % (inv) (RHS) National housing transactions (y) dependency on mortgage (x) 60,000 55,000 transactions (y) 50,000 45,000 40,000 35,000 30,000 25,000 y = x R² = ,000 15,000 10, Mortg age (%) (x) Source: ABS / RBA

3 2. The dependency of household goods sales on housing There are several factors that drive household goods sales, although we have found that housing (transactions involving existing housing stock, and lending finance associated with the purchase and construction of dwellings) hold some of the strongest relationships from a statistical standpoint. In terms of household goods sales data, there wasn t an overwhelming difference in correlation across various household goods categories, therefore the total was used. See example of New South Wales correlation matrix; HOUSING TRANSACTIONS NSW Construction NSW purchase new NSW purchase existing NSW TOTAL NSW furniture % 55.83% 69.21% 67.45% HH GOODS NSW electrical % 57.74% 72.04% 69.80% NSW hardware, building & garden % 56.20% 71.84% 69.49% NSW HH GOODS % 57.44% 72.10% 69.93% Charting of these variables also reveals a fairly strong positive relationship which was supported by our regression analysis implying that almost 70% of the movement in household goods sales is determined by residential transactions within the pool of existing housing stock. National housing transactions (#) vs HH Goods Sales ($m) 60,000 55,000 50,000 45,000 40,000 35,000 30,000 25,000 20,000 15,000 10,000 5, $5,000 $4,500 $4,000 $3,500 $3,000 $2,500 $2,000 $1,500 $1,000 $500 Source: ABS

4 goods sales (y) dependency on national housing transactions (x) $5,000 $4,500 $4,000 $3,500 HH goods sales ($m) (y) $3,000 $2,500 $2,000 $1,500 $1,000 y = x R² = $500 $0 Source: ABS 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 50,000 55,000 60,000 There was also a very strong positive relationship between lending finance (associated with the purchase and construction of dwellings), and household goods sales. In this case, regression analysis suggested that almost 98% of the movement in household goods sales is determined by housing related lending finance. goods sales (y) dependency on lending finance (x) $6,000 $5,000 goods sales ($m) (y) $4,000 $3,000 $2,000 y = x R² = $1,000 $- $- $5,000,000 $10,000,000 $15,000,000 $20,000,000 $25,000,000 Lending finance (purchase and construction) ($m) (x) Source: ABS We note there are other factors that could reasonably impact the sale of household goods such as; house prices, dwelling approvals, and population growth. However, while these factors make logical sense, the relationships are not statistically significant in the manner demonstrated by housing. For example, the correlation between household goods sales and dwelling approvals is only 60.37%, while the R2 is only 36.45%.

5 goods sales vs Private dwelling approvals (annual totals) 240,000 $55, ,000 $50, , ,000 $45, ,000 $40, , ,000 $35, , Total private dwelling approvals (#) approvals ave ('06-'13) approvals ave ('14-'17) National HH goods sales ($m) (RHS) $30,000 Source: ABS goods sales (y) dependency on dwelling approvals (x) 5,000 4,500 4,000 3,500 goods sales ($m) (y) 3,000 2,500 2,000 1,500 y = x R² = , ,500 10,000 12,500 15,000 17,500 20,000 22,500 Dwelling approvals (#) (x) Source: ABS

6 3. The dependency of rents on household goods sales Quarterly gross face rental series dating back to 1994, paired with household goods sales revealed one of the strongest statistical relationships. Charting and regression analysis revealed a strong positive relationship between these two variables implying that almost 96% of the movement in gross face rents was attributable to fluctuations in total household goods sales. National HH goods sales ($m) vs National ave gross face rent ($/sqm/pa) $5,000 $4,500 $4,000 $3,500 $3,000 $2,500 $2,000 $1,500 $330 $310 $290 $270 $250 $230 $210 $190 $170 $1, $150 Source: ABS/Colliers Research National ave rent ($/sqm/pa) (LHS) National HH goods sales ($m) (RHS) National rent (y) dependency on household goods sales (x) $330 $310 National rents (y) ($/pa/sqm) $290 $270 $250 $230 $210 $190 y = x R² = $170 $150 1,250 1,500 1,750 2,000 2,250 2,500 2,750 3,000 3,250 3,500 3,750 4,000 4,250 4,500 4,750 goods sales (x) ($m) Source: ABS/Colliers Research

7 What about the direct relationship between mortgage and rents? While the discussion so far steps us through the relationship between mortgage and rents, a direct relationship is harder to prove. Regression analysis between the two reveals quite a weak direct relationship, implying just 28% of rent movements are affected by mortgage. Charting the quarterly data shows a resemblance of a relationship, but the relative volatility of the mortgage compared to the smoother, less volatile rental series explains this weakness. rents ($/sqm/pa) vs (qtrly) $330 $310 $290 $270 $250 $230 $210 $190 $ $ National ave rent ($/sqm/pa) (LHS) Own occ var % (RHS) Own occ 3Y % (RHS) Source: RBA/Colliers Research rents (y) direct dependency on mortgage (x) $330 $310 $290 rents (y) ($/sqm/pa)y) $270 $250 $230 $210 y = x R² = $190 $170 $ Owner occ var mortgage rate (x) Source: RBA/Colliers Research

8 Would mortgage ever move high enough to impact rents? So yes, we have established that an inverse relationship exists between mortgage and rents, but how high would mortgage need to climb before housing slowed convincingly, setting off a series of events that would, according to our thesis, lead to lower rents? Taking a closer look at the following charts can assist in answering this question. National housing transactions (y) dependency on mortgage (x) 60,000 55,000 transactions (y) 50,000 45,000 40,000 35,000 30,000 25,000 y = x R² = ,000 15,000 10, Mortg age (%) (x) Source: ABS / RBA Lending finance (y) dependency on mortgage (x) $25,000,000 Lending finance ($m) (y) $20,000,000 $15,000,000 $10,000,000 y = -1E+06x + 2E+07 R² = $5,000,000 $ Mortg age (%) (x) Source: ABS / RBA It appears that a statistical relationship only starts to convincingly take hold after standard mortgage (as tracked by the RBA) move higher than 9 per cent, far higher than current levels (5.2 per cent according to RBA data). Until these more extreme levels are reached, housing transaction volumes and housing related financing deals appear to be somewhat inelastic, hovering around long-term averages. In this respect, the relationship is far less certain when standard mortgage as documented by the RBA remain within the range of 5-9%.

9 Cash rate target forecast (%) Source: Oxford Economics Even when we factor in the possibility of a higher RBA cash rate, the likelihood of these standard mortgage breaking through 9 per cent appears remote. Oxford economics are currently forecasting a move higher in the cash rate from late 2019, eventually reaching 3.25% by 4Q25. At this elevated level, corresponding standard mortgage, have historically only reached a range of %, far lower than the 9% tipping point, highlighted within our analysis. vs RBA cash rate (%) Mar-18 Jun-18 Sep-18 Dec-18 Mar-19 Jun-19 Sep-19 Dec-19 Mar-20 Jun-20 Sep-20 Dec-20 Mar-21 Jun-21 Sep-21 Dec-21 Mar-22 Jun-22 Sep-22 Dec-22 Mar-23 Jun-23 Sep-23 Dec-23 Mar-24 Jun-24 Sep-24 Dec-24 Mar-25 Jun-25 Sep-25 Dec-25 Source: RBA RBA cash rate Own occ var Own occ 3Y RBA target forecast Tipping point

10 Conclusion Our analysis demonst that it is possible to create a statistical relationship between the national housing market and the sector. While a direct relationship between mortgage and rents appears to be weak on face value, we can connect the dots through a series of logical relationships (mortgage, housing, household and rents), that are indeed statistically significant. At first glance, the establishment of such a relationship between mortgage and rents paints a subdued outlook for the sector, particularly in the current economic climate where the consensus outlook for global interest is shifting higher. However, upon further analysis we highlight that even though an inverse relationship exists, mortgage would need to move meaningfully higher before rents commence a convincing shift lower. In fact, standard mortgage (as tracked by the RBA) would need to lift from current levels of 5.2 per cent to above 9 per cent (>380bps), which appears unlikely, even after factoring in incremental hikes to the RBA cash rate from 1.5 per cent currently to 3.25 per cent by Colliers International does not give any warranty in relation to the accuracy of the information contained in this report. If you intend to rely upon the information contained herein, you must take note that the information, figures and projections have been provided by various sources and have not been verified by us. We have no belief one way or the other in relation to the accuracy of such information, figures and projections. Colliers International will not be liable for any loss or damage resulting from any statement, figure, calculation or any other information that you rely upon that is contained in the material. Colliers International Accelerating success.

Outlook for Australian Property Markets Brisbane

Outlook for Australian Property Markets 2009-2011 Brisbane Update August 2009 Outlook for Australian Property Markets 2009-2011 Brisbane Residential Update August 2009 Population growth continues to surge