The residential property market follows a cycle has lasting about 16 years, so on that reckoning we are about to re-live 2004. The cycle certainly plays out at the macro level, favouring different regions and cities depending on the cyclical phase, but what may go un-noticed without quantitative analysis is that the same cycle drives the relative performance at a much more granular level. Identifying and the current cyclical position and projecting the cycle forward to an investment horizon makes for more informed investment decision-making than simple intuitive approaches based on fundamental value or price momentum.
The driving force for the cycle is the ripple effect, with the epicentre of the ripples at the vertex of the property price pyramid: Prime Central London. Ripples are much more to do with price (£/m2) than with location, although the two are of course somewhat related. To illustrate these findings we will briefly explore the data in a very specific way, following paths through metro areas which traverse a range of prices. Each trail has been selected not so much for its natural beauty, instead to steadily descend in price over as many postcode sectors as possible.
Logically we start with Prime Central London, the veritable ground zero for each cycle, where handily for our purposes we find within the single postcode area of Southwest London a very broad spectrum of prices. We’ll follow a trail from the elegant mansions of Knightsbridge to the more affordable suburbs bordering Croydon which illustrates not only a fairly steep price descent within a few kilometres, but also the diversity of capital gain through the cycle. This is of particular interest given that we are now - after a long breather in Prime Central London - just past a turning point in the cycle.
The tabulated data is the result of revaluing to the present every property which has both sold since 1995 and received an Energy Performance Certificate, probably post 2008 when they became mandatory. Since we have sector-level indices monthly, we can do this for any point in time between now and 1995. In the tabulation we see the current snapshot with price data as of end October 2019.
| rank | Sector | Price/m2 | Total | Property average | |||
|---|---|---|---|---|---|---|---|
| Units | Present value (B) | m2 (000) | Price (M) | m2 | |||
| 1 | SW1X 0 | 23,700 | 580 | 1.8 | 77 | 3.11 | 130 |
| 2 | SW3 2 | 21,900 | 850 | 2.2 | 101 | 2.60 | 120 |
| 3 | SW3 4 | 20,500 | 850 | 2.2 | 109 | 2.63 | 130 |
| 4 | SW3 3 | 19,400 | 1330 | 1.5 | 77 | 1.13 | 60 |
| 5 | SW7 2 | 19,300 | 310 | 1.0 | 49 | 3.07 | 160 |
| 6 | SW7 3 | 18,600 | 760 | 1.5 | 82 | 2.02 | 110 |
| 7 | SW7 5 | 17,700 | 1010 | 1.9 | 107 | 1.86 | 110 |
| 8 | SW7 4 | 16,300 | 1120 | 1.6 | 96 | 1.40 | 90 |
| 9 | SW5 0 | 15,800 | 1340 | 2.1 | 131 | 1.54 | 100 |
| 10 | SW5 9 | 12,700 | 1220 | 1.1 | 90 | 0.94 | 70 |
| 11 | SW6 1 | 11,900 | 1080 | 1.1 | 90 | 0.99 | 80 |
| 12 | SW6 4 | 11,800 | 1170 | 1.5 | 124 | 1.26 | 110 |
| 13 | SW6 5 | 11,100 | 1130 | 1.3 | 121 | 1.19 | 110 |
| 14 | SW6 6 | 10,700 | 1610 | 1.8 | 168 | 1.12 | 100 |
| 15 | SW15 1 | 9,600 | 1700 | 1.7 | 180 | 1.02 | 110 |
| 16 | SW15 6 | 8,900 | 2060 | 1.7 | 193 | 0.84 | 90 |
| 17 | SW15 2 | 8,900 | 2770 | 2.1 | 230 | 0.74 | 80 |
| 18 | SW18 5 | 8,700 | 2480 | 2.0 | 235 | 0.82 | 100 |
| 19 | SW19 8 | 8,600 | 2810 | 2.1 | 243 | 0.74 | 90 |
| 20 | SW18 4 | 8,200 | 2430 | 1.4 | 166 | 0.56 | 70 |
| 21 | SW17 0 | 7,400 | 2280 | 1.3 | 171 | 0.55 | 70 |
| 22 | SW17 9 | 7,000 | 2190 | 1.2 | 169 | 0.54 | 80 |
| 23 | SW16 6 | 6,400 | 2130 | 1.2 | 182 | 0.55 | 90 |
| 24 | SW16 2 | 6,200 | 2340 | 1.1 | 185 | 0.49 | 80 |
| 25 | SW16 3 | 5,500 | 1050 | 0.7 | 121 | 0.63 | 110 |
| 26 | SW16 4 | 4,900 | 1550 | 0.7 | 142 | 0.45 | 90 |
These are not trivial moves, and it behooves the investor or homeowner to quantify them as accurately as possible when making investment decisions, and to be alert for turning points.
Putting this together with the tabulated data, consider an investor buying in Knightsbridge at the highs in 2014 - ‘to beat the mansion tax’ - at the average price of over £3M back then. They subsequently underperformed a diversified investor in Norbury by 30%, at a cost of £1M. Had they looked further afield for example to Manchester, the opportunity cost would have been more stark. Saving tax sometimes drives poor decision-making.
Next consider Birmingham, where the top-priced sectors nearly match the bottom-priced sectors in our London trail. Here the largest divergence occurred 1995 to 2001, when the leafy suburbs to the south of Solihull (now £4400/m2) and beyond more than doubled in price. Over the same period the lower-priced inner districts such as Saltley (now £1400/m2) returned less than 20%. The subsequent 8-year convergence largely reversed this, as the cheaper areas returned well over 100% and the same southern suburbs some 30-40%. Fast-forward to the end of 2018 and Birmingham has entered a new convergence phase.
| rank | Sector | Price/m2 | Total | Property average | |||
|---|---|---|---|---|---|---|---|
| Units | Present value (B) | m2 (000) | Price (M) | m2 | |||
| 1 | B93 0 | 4,500 | 740 | 0.4 | 83 | 0.50 | 110 |
| 2 | B93 9 | 4,400 | 1190 | 0.6 | 146 | 0.54 | 120 |
| 3 | B91 3 | 4,100 | 1870 | 0.9 | 214 | 0.47 | 110 |
| 4 | B91 1 | 3,900 | 1800 | 1.0 | 248 | 0.53 | 140 |
| 5 | B91 2 | 3,500 | 2170 | 0.8 | 219 | 0.35 | 100 |
| 6 | B92 7 | 3,100 | 2060 | 0.6 | 197 | 0.29 | 100 |
| 7 | B92 8 | 3,000 | 1970 | 0.5 | 173 | 0.27 | 90 |
| 8 | B92 9 | 3,000 | 1300 | 0.3 | 109 | 0.25 | 80 |
| 9 | B26 3 | 2,400 | 1580 | 0.3 | 126 | 0.20 | 80 |
| 10 | B26 2 | 2,200 | 1590 | 0.3 | 132 | 0.18 | 80 |
| 11 | B33 0 | 2,000 | 1180 | 0.2 | 91 | 0.15 | 80 |
| 12 | B37 6 | 2,000 | 1220 | 0.2 | 101 | 0.16 | 80 |
| 13 | B34 7 | 2,000 | 1070 | 0.2 | 86 | 0.16 | 80 |
| 14 | B33 9 | 1,900 | 1360 | 0.2 | 104 | 0.15 | 80 |
| 15 | B8 2 | 1,600 | 1810 | 0.3 | 168 | 0.15 | 90 |
| 16 | B8 3 | 1,500 | 1260 | 0.2 | 115 | 0.14 | 90 |
| 17 | B8 1 | 1,400 | 780 | 0.1 | 71 | 0.13 | 90 |
To go further down the price scale is to head either north or west. Consider Liverpool, which has a wide spectrum of house prices. In its strong convergence phase late 2001 to mid 2005 the bottom-priced sectors towards Bootle (now £700/m2) returned over 150%, while higher-priced southeastern sectors (now £3000/m2) returned nearly 100%. This dizzy frenzy was followed by a long hangover, so in the divergence from 2005 to 2017 the very highest returns were clocked up in the nicer parts, but even here did not reach 20%. Up toward Bootle, you lost over 20% over this nearly 12-year period. Since 2017 convergence has resurfaced and the cheapest areas have returned over 5% over the last year.
| rank | Sector | Price/m2 | Total | Property average | |||
|---|---|---|---|---|---|---|---|
| Units | Present value (B) | m2 (000) | Price (M) | m2 | |||
| 1 | L18 2 | 3,000 | 300 | 0.1 | 41 | 0.40 | 140 |
| 2 | L15 6 | 2,500 | 400 | 0.1 | 45 | 0.29 | 110 |
| 3 | L15 7 | 2,200 | 340 | 0.1 | 32 | 0.21 | 90 |
| 4 | L14 3 | 2,200 | 290 | 0.1 | 27 | 0.21 | 100 |
| 5 | L13 4 | 1,500 | 340 | 0.0 | 29 | 0.13 | 90 |
| 6 | L13 5 | 1,300 | 740 | 0.1 | 59 | 0.10 | 80 |
| 7 | L13 6 | 1,300 | 450 | 0.1 | 44 | 0.13 | 100 |
| 8 | L13 3 | 1,200 | 510 | 0.1 | 46 | 0.11 | 90 |
| 9 | L6 8 | 1,200 | 260 | 0.0 | 30 | 0.14 | 120 |
| 10 | L13 7 | 1,100 | 580 | 0.1 | 58 | 0.11 | 100 |
| 11 | L6 4 | 1,100 | 870 | 0.1 | 69 | 0.09 | 80 |
| 12 | L13 8 | 1,000 | 450 | 0.0 | 36 | 0.08 | 80 |
| 13 | L6 0 | 900 | 510 | 0.0 | 42 | 0.08 | 80 |
| 14 | L4 2 | 900 | 1060 | 0.1 | 91 | 0.07 | 90 |
| 15 | L4 5 | 800 | 1170 | 0.1 | 97 | 0.07 | 80 |
| 16 | L4 3 | 800 | 510 | 0.0 | 45 | 0.07 | 90 |
| 17 | L20 2 | 700 | 380 | 0.0 | 35 | 0.06 | 90 |
A hasty analysis would conclude that each city has its own cycle. Such a conclusion would miss a very important finding – that the market is driven by national factors, and the response of each area, district and sector is primarily driven by its price, or more precisely its £/m2. The only significant complication arises from differential sensitivities to the market as a whole, sometimes called the beta. The evidence for these sweeping statements does not however fit into this brief blog post and will have to await another, nevertheless the results presented here do support the assertion.
Turning now to the actionable points, the usefulness of this depends on whether you are a home-buyer, investor, or developer. On a cyclical basis, hit the road to Wigan for the tail end of this ripple, or perhaps Chelsea for the new wave. To take just one use case, consider the footloose London home-buyer - say the young grad who is not yet planning to grow a family. They should without hesitation head toward the upmarket end of their available price range. Now’s not time for gardening in the suburbs or the commuter belt, instead keep it posh, forego the spare bedroom, get a studio.
At the most basic level, we need to know whether we are in convergence or divergence, how long to the next turning point, and a probable dispersion of performance before we get there. Starting from repeat sales indices which optimally track not just a sample of transactions but the entire population, we have a comprehensive dataset which supports analysis of return and risk at each level of aggregation. Unlike security markets residential property shows extremely regular behaviour and long cycles, presenting significant opportunities to outperform through national and local choices informed by quantitative tools.
Giles Heywood
anest.uk
price paid last transaction: 2019-12-31
report run: 2020-01-31
This is a revised and updated version of the post to incorporate November data and extend to 2 further areas - it is substantially the same as the earlier post for SW