Dublin’s fifth office development boom reached its peak during 2009 marked by the completion of over 172,000 sq.m. of space, a 4 per cent increase on the previous year. In fact, this was the third highest recorded quantity of space to be developed in a single year, outstripped only by 2001 (263,870 sq.m.) and 2002 (177,800 sq.m.).


More than half of that new space was located in suburbia.

However, it is clear that the figure for 2010 will be significantly lower than for 2009, probably at around 100,000 sq.m., a reduction of some 40 per cent on last year, marking the end of the boom.

An interesting feature of recent developments was the increasing scale of individual schemes. Fifteen developments completed between early 2007 and late 2009 each accommodated over 10,000 sq.m. of floorspace. Together, they added over 260,000 sq.m. to the city’s office stock, a figure which equals the total of all the 85 developments completed between 1960 and 1972.

This increasing scale of schemes is reflected in the tendency for the number of development schemes to diminish over recent years. The figure shows that while the number of schemes reaching completion has reduced during the 2000s, the quantity of floorspace being developed remained high.

Thus, the average size of development due for completion in 2010 is 11,200 sq.m. compared to 2,513 sq.m. in 2001, a figure which differs little from the average development size in 1981 (2,364 sq.m.) or 1991 (2,709 sq.m.).

This increasing scale was in part a response to the increasing demand by certain companies for larger office units. For example, during 2006 there had been over 50 transactions each involving more than 1,000 sq.m. of office space. However, the increasing scale is also a clear indication of the growing confidence and the greater risks that developers and their financial backers were prepared to take as the market heated up. In 2009, only 14 transactions concerned take-up greater than 1,000 sq.m..

The development boom since 1995 has dramatically altered the geography of Dublin’s office stock. While in 1995 only 14 per cent of the stock was suburban, this had increased to 35 per cent by the end of 2009. Meanwhile, the proportion located in Dublin 2 fell from 54 per cent to 32 per cent.

Again, the willingness of developers and their financier to countenance locations outside the prime office core is a reflection of their confidence in the office boom in addition to the growing demand for space in more marginal locations.

In fact, the demand for space kept up relatively well through the period of the fourth and fifth development booms, reaching a peak in 2007 when 297,240 sq.m. was taken up.

Thereafter, take-up reduced dramatically and, in 2009, just 72,000 sq.m. was taken up, marking a 66 per cent reduction on the figure for 2008 and less than a quarter of the demand recorded in 2007.

Perhaps surprisingly, suburban take-up (at 42 per cent of the total) actually held up rather better than elsewhere, the figure exceeding what could be expected from the proportion of the total stock located there (36 per cent).

Nevertheless, the scale of completions in 2009 far outstripped demand.

Inevitably, vacancy levels rose. By the end of 2009, over 670,000 sq.m. of office space lay vacant. This figure was greater than the total stock of office space in 1980. It represents nine years of supply at 2009 take-up rates and, at prevailing occupancy rates, could accommodate some 33,500 staff.

Another way of conceptualising the scale of this vacancy is that it is equivalent to the amount of office space in 206 Liberty Halls.

About half of the total (53 per cent) lay in newly-completed buildings. Some 423,000 sq.m. (63 per cent of the total) had been vacant for more than twelve months.

Thus, from a historically very low vacancy rate which was recorded at 1.9 per cent in December 1999, the rate rose to 11.8 per cent by late 2001, then to 13.4 per cent in December 2007, to 15.7 per cent by the end of the following year and stood at 21.1 per cent in December 2009.

However, the citywide vacancy rate hides a number of interesting spatial variations.

Suburban vacancy accounts for half of the empty space in Dublin and the vacancy rate is suburbia is 30 per cent.

But the geographical variations in vacancy which typified recent years, with generally low inner-city rates and high rates of suburban vacancy, have also altered somewhat.

These have changed with geographical shifts in development activity (completions), the release of older space onto the market and changing patterns of take-up.

In the inner city, vacancy remains low in the International Financial Services Centre (IFSC). Indeed, this was the only office zone to register a reduction in its vacancy rate in 2009.

Vacancy in Dublin 2 rose to 15.5 per cent, to exceed the vacancy rate prevailing in the inner office fringe (Dublin 1, 7 and 8). This was occasioned by the release of older space and the completion of some large developments in the docklands.

Dublin 4 has also been typified by a rapid rise in its vacancy rate to 25 per cent, again due in large part by the completion of large new developments where 70 per cent of its vacant stock is located.

In suburbia, vacancy rates rose in all locations. Most noteworthy has been the increasing rate of vacancy in the north suburbs where 44,000 sq.m. was developed in 2008-9, the vacancy rate rising from 28 per cent at the end of 2007 to 46.5 per cent in December 2009.

A review of the location of floorspace that was vacant for over 12 months at the end of 2009 reveals the difficulty of finding occupiers for space in the western suburbs, in the inner-city fringe (Dublin 1, 7 and8) and for older space in Blackrock-Dun Laoghaire.

Indeed, a significant quantity of space in the western suburbs had been vacant for more than five years.

Andrew MacLaran

Savills published their Q2 2010, “Dublin Office Market in Minutes” report on August 3rd.  According to the report there is “782,500 sqm of vacant office stock” in Dublin, Q2 2010*.  Savills first quarter report details that overall vacancy rate in the city is 21%.   Of this, 23% is in Dublin 2 (c.180,000 sqm), 14% is in Dublin 4 (c.109,500 sqm), 21% is in Dublin 1,3,7,8 (c.164,000 sqm), and 19% in the West Suburbs (c.149,000 sqm).

With around 110,000 sqm of new office space due to be completed in 2010, it is difficult to believe that the city will be short of office space for a while, especially given the state of the economy and the number of companies either going out of business or downsizing.  It might be the case that landmark/headquarter buildings will come into short supply if they are taken up by large anchor tenants, as argued by Savills, but there is undeniably a substantial amount of available office space all round the city, including large developments, in prime locations, and much of it quality stock completed in the last few years (some that has never been occupied and some secondhand).

They conclude that “It is now clear that the office market has bottomed out and hopefully this will encourage “fence sitters” to progress their requirements before much of the better space is gone.”  There is clearly a massive amount of stock available at present in all parts of the city, including brand new and recently completed stock, providing would-be tenants with negotiating room.  If the market has bottomed out, it seems likely that growth will be very gradual as the economy starts to stabilize and claw its way back.

Perhaps then it is no bad thing that Savills report that “the supply of new stock to the market will cease in 2011 with no new developments due to be completed in 2011 or 2012” as it will allow for some market correction.  Take-up in Q2 2010 was 23,003 sqm through 28 deals (21,000 sqm in Q1 2010), and Savills estimate take up this year will be around 100,000 sqm (based on deals lined up but not signed).  (They do not estimate how much space will become vacant in the same period, but clearly some space will become empty as other space fills, especially if a company is simply relocating to take advantage of falling rents).  It may take some time then, at the present rate of take-up, to drastically reduce the vacant office supply in the city.

* I’ve updated this post.  I originally had an estimate of how many Liberty Halls this would translate into (I’ve now found three different total sqm estimates for the building, that differ quite substantially, so have removed until I can establish the correct size).  As an alternative comparison, MacLaran and Kelly report that at the end of 1990 there was 1,055,490 sqm of office space in Dublin (MacLaran, A. and Kelly, S. 2007. Urban property development, in Bartley, B. and Kitchin, R. (eds) Understanding Contemporary Ireland. Pluto Press, London).

A striking feature of the official NAMA documentation and subsequent commentary is its lack of attention to geographic or spatial considerations. Indeed, this tendency of ‘aspatial’ analysis has also been a key feature of banking and financial operations, where property assets, which exist in physical space and have a specific geographic location, are presented in numerical form (often aggregated) on loan books, on balance sheets and indeed as part of complex asset bundles, such as CDOs (collatoralised debt obligations). Such an understanding and presentation of data serves very much to conceal the geography of those assets.

During the boom, financial institutions massively expanded their total lending and, as we know, what characterises the Irish banking system (and crisis), is that much of the lending went into property, which has a very particular geography. And, given the general tendency of property developers and investors to engage in developments in increasingly riskier locations the longer the boom goes on, and the considerable property price inflation that occured in every region in Ireland, it was the overlooking of geographical considerations that facilitated further inflation of the asset bubble.

That push to boost lending and increase the profitability of banks seemed to pay little attention to broader indicators in the Irish economy but also failed to pay due cognisance to locational considerations. In an industry in which the three rubric for success have long been recognied as “location, location and location”, such overlooking of spatial considerations is nothing short of amazing.

The very viability of NAMA as a financial project is highly dependent on the geography of those property assets which underpin the loans which NAMA is purchasing. While certain well-located assets may indeed provide for a secure yield in the medium term, others, such as large sites lacking any urban zoning status located at the edge of small Irish towns and villages for which enormous prices were paid per hectare, are unlikely even in the very long term to prove to possess much value above that of farmland.

Approaching the ‘details’ of NAMA and applying sectoral and geographical scenarios
From the NAMA Draft Business Plan, one can ascertain the broad composition of the loans. Of the total amount of €77bn of loans being transferred to NAMA from banks’ asset sheets:
• €27.8bn (36%) relates to ‘land’
• €21.8bn (28%) relates to ‘development loans’
• €27.7bn (36%) relates to ‘associated loans’
Beyond that broad breakdown, there is little information on when the loans were granted, on the type of property to which the loans relate (residential, office, retail, industrial, etc.) and on the geography of the assets.

To date, the total extent of our knowledge of the ‘geography’ of those loans is provided only by a very broad break-down between international locations. But in a sector in which a matter of 40 or 50 metres distance between sites might mean the difference between success and failure in the case of urban redevelopment projects, particularly in office development markets, such crude delimitation is almost entirely unhelpful. Unfortunately, we know too little about the location of NAMA-related property for an accurate assessment to be made as to their likely real value.
Some €33.13bn of those loans relate to properties located in the Republic of Ireland. However, an appraisal of the real value of the underlying assets requires, in the case of loans against land, a review of the geographical location, zoning characteristics and planning status. For development loans, an evaluation needs to be undertaken of the sectoral status (residential, industrial, office, retail and other commercial) of schemes within a local context in which the current level of provision of such property is taken into account, together with considerations of existing vacancy rates, the projected scale and stage of development (completions, currently on-site or due to be undertaken), and the prevailing and projected scales of demand for such space.
As laid out in the draft NAMA business plan, NAMA will pay €54bn for loans with a book value of €77bn, representing a 30% “haircut” on the original loans’ value. While NAMA will pay 15% above the estimated current market value for the assets, it does so on the basis that the total reduction in average property prices in the boom-slump cycle is 47% (the peak having occured in early 2007 and the trough estimated to have occured in September 2009) and that the 15% premium can be recouped over the lifetime of NAMA through a rising property market.
While we do not have specific information on the geographical location of the assets that are to be transferred to NAMA, we can examine some recent trends in the property market and compare them to the assumptions underpinning NAMA.

The Agricultural Land-Price Spiral
What became clear during the boom years was that a very sharp increase in land prices was occuring, not just in and near urban areas but also in rural areas – agricultural land. According to SavillsHOK’s Irish Agricultural Land Research report in May 2007, Irish land values jumped from just under €10,000 per hectare in 1998 to over €58,400 per hectare in 2006, by far the highest price per hectare in Europe. This jump in part reflected the tendency for farmers to sell small plots of land for development purposes (one-off housing) and an increasing trend of wealthy individuals buying small farms for lifestyle purposes, particularly within commuting distance of Dublin. In many cases, the proceeds from land sales was re-invested in farmland, where farmers with land at the edge of towns sold them off during the boom times and bought farms elsewhere pushing up land prices even further and becoming even more out-of-kilter with the day-to-day economics of farming. With the onset of the property crash, agricultural land prices have and will continue to fall back sharply, with those greenfield sites at the edge of towns bought with redevelopment in mind, open to devaluations in the region of 90-95%.

Record Site Prices (see google map)
If we examine some of the record prices paid for redevelopment sites in urban areas at the height of the boom, and compare them to revaluations on other similar sites during the slump, it should give us some indication as to the real extent of price falls.
In late 2005, a number of significant site sales in the prestigious area of Ballsbridge in Dublin 4 attracted considerable attention. During this period, the developer Sean Dunne bought two sites, the former Jury’s Hotel and the former Berkeley Court Hotel, 7 acres in total for ca. €379M (€54.1M per acre). Another developer, Ray Grehan of Glenkerrin Homes bought a nearby 2.05 acre site, the former UCD Veterinary College for ca. €171.5M (an incredible €83.7M per acre). In 2006, a small site of just over a third of an acre went for almost €40M (€95M per acre). These were purchases that were made close to the height of the boom for potential redevelopment sites which had no planning permission. Even if the subsequently proposed mixed-use developments had been given planning permission and had they been completed at the height of the property boom, the margins for profitability were already stretched. So what scale of price devaluations could we expect in the property crash for these sites?
Take the example of a nearby 0.5 acre site on Herbert Road, Ballsbridge; this former Cablelink office-building was offered for sale in autumn 2007 for €30M, just after the peak of the property boom. Two years later, the asking price stood at €9.5M – a 68% decrease, and that’s just the asking price!

As already mentioned on this blog, the case of the Irish Glass Bottle site in Ringsend is instructive. This 24.9 acre site, purchased in early 2007 for ca. €413M (€16.6M per acre) by a consortium involving companies of Bernard McNamara and Derek Quinlan and the Dublin Docklands Development Authority (DDDA), was revalued in October 2009 at €60M. This represents a whopping 85% price devaluation, which is way above the “haircut” of 30% or the 47% average property price fall upon which NAMA’s calculations are based. In the case of the Irish Glass Bottle site, we are talking about a large centrally-located brownfield site with special planning status for mixed use (residential and enterprise) functions under the DDDA’s Master Plan, with reasonably good prospects of redevelopment in the medium-to-longer term. Yet, this site has retained only 15% of its market-peak value. This begs the question, if such well-located sites are experiencing such colossal price falls, then what are the prospects for greenfield sites at the edge of cities or at the edge of towns throughout Ireland?

Investigating Development Loans – The changing geography of Dublin’s office stock
Presumably, many of the Development Loans and Associated Loans in the NAMA business plan relate to loans for commercial property in Ireland. So, let’s examine some recent trends and indicators for the different commercial property sectors. Taking the Dublin office market as an example, research from the Centre for Urban and Regional Studies (CURS), TCD and SavillsHOK (MacLaran & O’Connell, 2007) has traced the changing volume and geography of office space in Dublin since 1995. Over the past 15 years, the volume of modern office stock expanded enormously. However, the geography of office development has also changed significantly, where proportionately, there has been a fall in the dominance of the prime office locations of Dublin 2 and 4, while suburban office locations have grown substantially. While in 1995, suburban office developments accounted for 14.6% of the total stock, by 2008, that figure had leapt to over 35%.

Rates of vacancy in the modern office stock vary considerably geographically and are particularly high in certain suburbs, having been especially high in west Dublin. The research indicates that since 2007, there have been high levels of activity in office construction in suburban areas, office space that has yet to reach the market. It is likely that some loans associated with office space in suburban areas will be transferred to NAMA, areas that even during the boom were experiencing relatively high vacancy levels.

Moreover, with the Minister for Justice Dermot Ahern recently banning the use of upward-only rent reviews for all new leases in commercial property from March 2010, this is likely to have a downward effect on the capital values that new office buildings are likely to achieve. While, this ban could be seen as a positive step for some business sectors (new tenants especially), it may have serious implications for NAMA (and simultaneously for Irish society) as it is tied to a rising property market. More specifically, many of NAMA’s underpinning calculations drew heavily on trends and expectations in commercial property sectors, which at that time were based on a system of upward-only rent reviews!

Residential Loans
It is likely that a good deal of the Development and Associated Loans relate to residential developments at various stages of construction (see discussion of ghost estates elsewhere on the blog). And, while it is clear that substantial reductions have occured in house prices generally, it is difficult to gauge the likely level of devaluation on residential loans that will transfer to NAMA. Where loans transferred to NAMA are found to involve housing developments that are at a stage of completion, we can draw on new house prices as an indicator.

Source: DoEHLG, Housing Statistics Bulletins, various years (Average house prices are derived from data supplied by the mortgage lending agencies on loans approved).

Based on the DoEHLG’s Housing Statistics Bulletins to July 2009, new house prices nationally fell by 26% since the peak of the boom (2007Q2). For Cork and Galway, the figure was 23%, for Limerick, 11%, with Waterford and other areas experiencing 26% and 25% price drops respectively. Notably, the fall was sharpest in Dublin where new house prices fell by 40% and there is little to indicate that the downward trend halted in September 2009. Unsurprisingly, sales activity is much reduced.
Notably, the type of ‘assets’ that will be transferred to NAMA will likely include some of those unfinished housing estates and apartment complexes that have become a very visible feature of the Irish landscape – in city suburbs, in inner-city neighbourhoods and at the edge of towns and villages countrywide. In these cases, the ratio of market value to loan value will have fallen dramatically, with the management and disposal of these ‘assets’ likely to prove difficult and expensive.

The property crash beyond NAMA
While this post has sought to develop further a geography of NAMA’s portfolio, raising some serious questions about the calculations and underpinning assumptions of NAMA in the process, it must also be remembered that there are other major elements of the property crash that will not be addressed by NAMA. The NAMA portfolio involves loans relating to ca.1,600 borrowers, with just 100 borrowers accounting for 50% of the portfolio. Land, Development and Associated Loans of less than €5M are excluded from NAMA (except those with EBS and Irish Nationwide); the number of these loans is unknown but their geography is likely to be predominantly rural in nature or rather, at the fringe of towns and villages throughout Ireland. Then, of course, there is that far more immediate concern for tens-of-thousands of households of mortgage repayment difficulties, negative equity and the threat of housing foreclosures in the not-too-distant future. In NAMA and the SPV, these households will find little comfort – but that’s another day’s blog.

Sinéad Kelly

<!–[if !mso]> <! st1\:*{behavior:url(#ieooui) } –>

The Changing Geography of Dublin’s Modern Office Stock 1995-2008

Source: CURS/SavillsHOK Office Database