Local Heroes: How we differentiate areas and why it matters for real estate performance

Commercial property returns are variable by geography. This dispersion is significantly influenced by the resilience and growth in income. In turn, income is influenced by the strength of the labour market at a local level. Understanding the local economic and demographic drivers of employment, and how they vary, is therefore critical to identifying areas of outperformance. In this paper we highlight what makes a difference at a local level and how we embed it into our decision making.

Why employment matters

Commercial property returns are divided into income return and capital growth. Over the long term, two thirds of returns have been derived from income.

Income return is driven by occupational markets; the demand for operational space for businesses and trading space for retailers. The higher the rate of occupier retention within buildings, the more stable the income return. The growth and resilience of income is therefore crucial for long term outperformance and investment success.

Many of the drivers of income can be found in the local economy. This is because economic output explains a significant amount of the difference in employment growth between areas. And differences in economic output can be explained by differences in productivity.

Employment growth is, therefore, strongly associated with the investment performance of property. Figure one below shows the extent to which this is the case. We divided local authorities which had property performance data into quartiles based on historic employment growth. Concurrent property performance, measured by total return, is clearly associated with employment growth. The magnitude of the differential is notable – a difference in performance of over 4% per year over the 10 years to 2016.

Figure one: Average return by employment growth location, 10 years


But what drives employment? A bit of theory

There is a wealth of academic resource focussed on explaining local economic differentials. We can only skim the surface in this article but a useful one-word shortcut is ‘education’.

This cause is championed by many, but one of the most well-known protagonists is Professor Edward Glaeser from Harvard. He has consistently found that education to be the best predictor of subsequent economic growth and this has profound implications for public policy. There is a conflict, he claims, between investing in “vanity infrastructure projects” versus education (at all levels). The implication being that a failure to invest in the latter is tantamount to a dereliction of economic stewardship.

To understand local differentials within the UK a short recap of history is required. Across the UK deindustrialisation and a failure to re-orientate subsequently has much to do with a significant 650% differential in long term jobs growth at the local level1. According to the Centre for Cities, areas can be described as “replicators” or “re-inventors”: either trying to do what they did before – uneconomically - or trying something new.

"The majority of recognised growth drivers hinge on the skills of both the population and the workforce. This is self-reinforcing"

In other words, 150 years ago the purpose of cities was low cost production taking advantage of low cost workers. Now success utilises workers’ knowledge. The majority of recognised growth drivers hinge on the skills of both the population and the workforce. This has been found to be self-reinforcing: the higher skilled an area is the more likely it is to remain so.

Figure two: Output and employment growth, major UK Cities, 1990-2017


Besides education, other factors identified by academia include infrastructure; entrepreneurship (a thriving business climate); mobile labour (better educated labour tends to be more mobile) and finally – and less measurable – a sound political economy with a level of fiscal autonomy. As mentioned above these factors explain productivity, productivity drives output, and output drives employment. There is a clear relationship between the two as illustrated in figure two above.

Figure 3: Domains of growth: identified factors influencing local area success


How do we approach it?

We have combined theory with experience and created a new framework with which we can guide investment. For every local area in the UK we have measured factors known to influence employment growth and expressed them as a percentile relative to their peers. We have also measured factors which we suspect to be influential, but where empirical causation cannot be demonstrated. The indicators can be grouped into domains, as illustrated in figure three above. This first stage, expressed as a scorecard for each of the UK’s 390 local authorities and 66 Primary Urban Areas, provides a useful empirical resource for the understanding and comparison of local areas.

The next stage is prediction. By regressing local characteristics with employment growth we can identify what combination of factors explains most of the differences between employment growth at a local level. This relationship, based on what has happened over the last 15 years, can then be applied to present day economic and demographic characteristics to determine a forward looking assessment of an area’s long term prospects.

There are aspects of local performance such a method cannot capture empirically, the key ones being local policy or the effectiveness of governance. In addition, we cannot capture consistently the pipeline of major developments in the planning system which may make a difference to an area’s prospects. However, this can be an overlay based on experience.

"This granular analysis can provide insights for areas beyond London and “the big 6” markets. It can also identify areas of mispricing"

The map overleaf shows all local authorities by percentile ranking, based on expected employment growth. This granular analysis can provide insights for areas beyond London and ‘the big 6’ markets. It can also identify areas of mispricing – either to the upside or downside. Areas like Cambridge, Oxford and Bristol score highly but so do smaller cities such as Watford and Milton Keynes which may otherwise fall below investors’ radars. Figure four below highlights some of the best prospects within four tiers based on population size.

Figure four: Top locations by tier*


The obvious question is: does this work? We have backtested the explanatory power of scores attributed to local authorities and subsequent property performance. Figure five shows the average annual total return from 2007 to 2016 based on locational scores as at 2006. The results show a compelling property performance gap of 3.7% p.a between areas.

Figure five: Scores as at 2006 with subsequent property performance


There is an important point to be made about cities and their constituent local authorities. For convenience it is tempting to use a Primary Urban Area (PUA) measure to capture the attributes of a City. Manchester PUA is a combination of 10 local authorities, for instance, and this aggregation seems intuitive. There are 66 PUAs across the UK, as defined by Centre for Cities, many of which combine more than one local authority. However, aggregated PUA scores can be dragged lower by their weaker constituents. This is the case in Manchester, where attributes of the City centre and Trafford are obscured by weaker fundamentals elsewhere. In Birmingham, the City and Solihull are strong in isolation but Birmingham drifts toward average when other areas are included. We therefore apply caution in the analysis of City level results and look for the strongest constituents to guide our decisions.

Map: UK local authorities by employment growth expectations


Talent leakage: Addressing a missing ingredient

As previously articulated, skills matter and influence a variety of our chosen variables. However, there are instances where an area has a high proportion of students and a low proportion of residents with a degree. In our view this is evidence of market failure: graduates are not being retained and we can call this ‘talent leakage’. This is most likely due to job prospects but will also be influenced by real estate – the physical ability of an area to house and attract the knowledge intensive companies to employ the area’s graduates.

We illustrate this in Figure six. The quadrants show areas relative to average. Cambridge and Oxford, as expected, fall within the top right quadrant – exemplified by areas with a higher than average proportion of students and resident education. However, it is the locations in the bottom right which are most intriguing: areas with a higher than average presence of students but a below average resident education. This includes notable university towns like Nottingham, Preston and Leicester. Direct investment into knowledge intensive sectors could, according to theory, lead to a greater retention of graduates and better long term prospects for overall employment growth. This will reinforce property performance.

Newcastle has an above average proportion of students but an average (albeit improving) skill base. It has compelling attributes overall but its office market, for instance, has underperformed by -5% p.a. over the last 10 years. Legal & General has invested £350 million into the City via a joint venture with Newcastle City Council and Newcastle University in the development of Newcastle Helix. This arguably creates a missing ingredient to local performance by encouraging – and enabling - the retention of talented graduates within the area and fostering a knowledge economy via provision of workspace. This will not be a panacea, of course, but it should narrow the performance gap.

Cambridge, meanwhile, is home to the world-renowned university. It has a student population and resident education in the 100th percentile with some very compelling additional attributes which place it in the 96th percentile overall.

Its property market has outperformed the national benchmark by 3% p.a. over the last 10 years. LGIM Real Assets manage in excess of £250 million of assets in Cambridge, but further acquisitions have been challenged by a rapid compression in yields bringing valuations in line with areas of London and provoking questions over value. Our analysis shows that long term investors in Cambridge are likely to enjoy outperformance given the market’s fundamental advantage in attributes known to be accretive to employment growth. Thus, a lower yield to its peer group could be justified.

Figure six: Comparing the dominance of a university with residents’ education


Conclusion and application

A detailed approach to understanding the drivers of employment growth and the application of this to determine future expectations can create a competitive advantage in direct property investment. Local areas most likely to outperform over the long term can be identified, and mispricing to the upside or downside can be exploited. It can also identify longer term regeneration need by focussing on areas with a key ‘missing ingredient’.

Investors’ motivations vary. Investors will have different costs of capital, risk tolerance and duration. This framework can help link these different pots of capital to appropriate destinations by, for instance, identifying ‘proven performers’ which will have a strong track record but are more likely to be keenly priced or “regeneration need” where significant, patient, capital may need to be deployed but where there is conviction on long term prospects. We summarise this in the simplified diagram figure seven below.

In summary, this framework is designed to guide investment decisions and provide structure to challenge proposed investments. It is flexible depending on investment horizons and risk tolerance. Most importantly, it is way of understanding local area differences and how and why these matter to property market performance.

Figure seven: Investment approach according to risk tolerance and horizon


Please get in touch with your comments or suggestions for future Market Insights topics:

  • 020 3124 2803
  • contactrealassets@lgim.com


Fig 1: MSCI, IPD Local Markets, NOMIS, LGIM RA calculations
Fig 2: Oxford Economics, adapted from Divergent Cities in Post Industrial Britain, Martin et al, 2015
Fig 3: GIM Real Assets
Fig 4: Nomis, LGIM RA calculations
Fig 5: MSCI, IPD Local Markets, NOMIS, LGIM RA calculations
Map: Moodys, Thomson Reuters
Fig 6: NOMIS, UK Census, LGIM Real Assets calculations
Fig 7: LGIM Real Assets