GRESB Performance Indicators – How?

Wednesday, June 17, 2015

As a part of the blog series on GRESB’s Performance Indicators - ‘Why, how and what’, I continue with ‘how?.’ Before I delve into some of the details of reporting performance indicators to GRESB, and look at how the approach we take is aligned with or different from leading disclosure protocols and initiatives, I would like to point out that we are focusing strongly on alignment with these protocols and initiatives (to the extent possible and applicable).

Key components

Let’s start with the key elements of GRESB’s approach to assessing performance indicator data. As stated in the previous blog posts on performance indicators, the overarching approach GRESB has taken stems from the starting point of transparency, measure to improve and focus on reduction: (1) report absolute data, (2) benchmark data coverage per property type and (3) benchmark like-for-like change per property type.

Reporting ‘what you’re able to collect’ in terms of absolute data is one thing, but knowing for what proportion of your portfolio or, even more complicated, the proportion of the assets within your portfolio is included, is a different ballgame. Until 2013, GRESB participants were able to self-report their portfolio coverage for energy consumption, GHG emissions, water consumption and waste data. It was challenging for GRESB to validate the data reported. In addition, we received quite some feedback on the structure of the performance indicator tables and the difficulty to report data in line with internal approaches used to collect and report data. On top of that, the complexity of collecting tenant data in several cases and especially in specific lease structures proved to be a challenge as well. A good example of this are FRI leases, common in the UK, which in many cases provide a single tenant with ‘operational control’ for a very long period of time and the landlord therefore does not have the ability to introduce and implement operating and/or environmental policies and measures. This is why GRESB has defined ‘Indirectly Managed Assets’.

We decided to rethink the structure of the performance indicator tables for the 2014 GRESB Survey to (1) enable participants to report their performance indicator data in line with the way the data is collected and structured initially and thus increase ease of reporting; (2) be able to have a better understanding of the actual data coverage and improve data quality; (3) separate data and thus data coverage for Indirectly Managed Assets from Managed Assets and benchmark these separately. This resulted in the current approach we have taken since 2014, and we don’t intend to make structural changes in 2016.

Flexibility of the tables for energy and water consumption

How should the performance indicator tables be interpreted and what basic knowledge do you need to get started? Be prepared for a fairly technical explanation. First of all, data is reported per property type. Second, you should make a distinction between Managed and Indirectly Managed Assets. Note that the distinction between the two should only be made at asset level and is unrelated to the ability to collect data within certain assets. It is related to GRESB’s definition of Indirectly Managed Assets and therefore purely focused on the landlord-tenant relationship and operational control with regards to the ability to introduce and implement operating and/or environmental policies and measures. Third, not all rows in the tables are mandatory; you should only complete the rows that are relevant for your portfolio. So, if you would only have ‘Whole Building’ data, only complete the applicable rows in that section. However, if you would report on ‘Common Areas,’ we also expect data to be completed in the ‘Tenant Space’ section, to make sure data for the whole building is reported. In the case no consumption data is available, at the minimum Maximum Coverage should reflect the floor area for which there is supply. If no supply of a particular energy type is available throughout your portfolio, simply leave those rows blank.


PIs - How


Data Coverage and Maximum Coverage

The breakdown of the tables for energy and water consumption (2015 Survey Guidance) also gives GRESB participants the opportunity to get a better understanding of their ability to collect data. It will show to what extent data is missing in particular areas of the building throughout the portfolio. GRESB acknowledges that it is more complicated to collect tenant data, especially for certain property types like residential, retail and industrial. However, from the starting point ‘to measure in order to improve’ tenant spaces and therefore tenant data should ultimately be included as well.

How does GRESB actually calculate data coverage? This is based on Data Coverage and Maximum Coverage, both to be reported by participants. The latter is the area for which there is supply (supply per energy type or water supply). Data Coverage is the area for which participants are actually able to report data. This might sound fairly simple, but completing this per asset and eventually for a whole portfolio has proven to be more complex. It starts with the inability of participants to report the actual size of their assets, or more precisely the common areas, since in many cases only lettable floor area is available. As an engineer, I must say I was surprised when I discovered how often this is in fact the case. It would be great to eventually see implementation of the International Property Measures Standards – having knowledge of measurements of assets is common sense in the financial world. Knowing which areas within a building are ‘covered’ by a sub-meter appears to be another challenge. Next to that, exceptions within portfolios, or extreme cases and complex portfolios increase the burden of reporting performance indicator data in general, and more specifically of data coverage. Nevertheless, the current structure has improved GRESB’s ability to get a better understanding of actual data coverage and benchmark the data per property type as well.


PIs - How2


Like-for-like Change

As Matthew Tippett stated in his first blog post in this series, like-for-like is the simplest way to check if performance is getting better or worse due to management activity. This comparison provides a better understanding of the change in consumption of a portfolio that has been consistently in operation during the full two consecutive periods of reporting, with the purpose of understanding whether efficiency measures implemented throughout the portfolio actually had the impact they intended to have – a straightforward and effective approach.

Assets that were purchased or sold during the reporting period or the year before should be excluded from like-for-like. GRESB does allow for estimates when landlord-obtained utility consumption data is partially unavailable or unreliable for an asset. Estimation allows complete annual data to be calculated for an asset where data is partially missing or unreliable – though this should not be used as a substitute for gathering complete and accurate data (2015 Survey Guidance).

There are a couple of downsides of a like-for-like comparison. The methodology does not account for a several of elements:

Vacancy. Increased vacancy can have a positive impact on the calculated like-for-like change or the other way around.

Improvement potential of the portfolio. In case efficient (low improvement potential) assets are sold and less efficient assets (high improvement potential) are purchased, the following will happen: these assets will initially both be excluded from the like-for-like comparison. Prior to the sale the improvement potential was fairly low and after the sale and purchase of new assets, the improvement potential became much higher. Within a portfolio this can obviously be explained, but within a benchmark this is much more complicated.

Weather conditions. Cold winters and/or hot summers are not taken into account, this holds for local differences within a portfolio as well as in the benchmark.

These elements are part of the main arguments why to move forward and include intensities in the GRESB assessment. This enables participants to report change over longer time frames and normalize for relevant indicators. This is currently one of the GRESB development topics, as there is no clear guideline available of how to calculate intensities globally, it is challenging to come up with a global approach within our assessment.

I’m curious to see Matthew Tippett’s addition on “How to report Performance Indicators” from a market perspective and will follow up with “what” in due course!