GREs versus the Private Sector in the UAE

Every once in a while I decide to torture myself and rummage through the IMF’s databases looking for interesting research and analysis. When I found a Selected Issues UAE Country Report by the IMF, I thought I’d try my luck.

The report, published this month, begins by looking at government-related entities (GREs), which is anything that the government owns shares in. It is important to note that the IMF repeatedly warns that it does not have all information on all GREs. It looks at about 60 companies, although one should bear in mind that the government holdings in some are too small to have any influence.

One of the early IMF comparisons that is striking is the return on assets (ROA) of the non-financial corporate sector across GCC countries over the period from 2007 to 2014. The UAE at 8.1% a year is higher only than Kuwait. Saudi Arabia at 9.6 per cent is about a fifth higher and Oman’s 13.3% is more than three-fifths higher. How then are we the commercial hub of the GCC?

In a continuing discussion of whether the government is crowding out the domestic non-financial private sector and the implications that has, the IMF calculates the representative GRE ROA over the period at 2.5% a year. The private sector, consistent with all theory and research, at an 8.1% ROA has outperformed the government by 225% a year. Now there are plenty of arguments to be made that the IMF’s computation of the GRE’s ROA is not completely accurate, but one has to draw the line at believing that the IMF could be wrong by 225%.

The IMF is diplomatic in its conclusion – leave business to businessmen.

There is some good news. The total GRE debt as a percentage of GDP has dropped strongly in Abu Dhabi and to a lesser extent in Dubai. Importantly for Abu Dhabi, the percentage of GRE debt that is loans as opposed to bonds dropped to 33% from 68%. This is important, as it allows private companies more room to borrow.

Abu Dhabi is showing a clear corporate governance push to stop GREs from crowding out the private sector in the loan markets.

For some fresh news, Mubadala is buying 20% of Investcorp. Why? The Ipic-Mubadala merger was just announced; that’s a lot of work. And Mubadala already owns 10% of Carlyle, arguably the top private equity firm in the world.

This is probably forward thinking by the team at Mubadala, a global investor. Merging with Ipic means also acquiring Aabar. Investcorp, with its long history, blue chip name and seasoned team, could make a perfect match for Aabar. A reverse takeover would make Aabar liquid again through an equity swap for Investcorp shares. As a result, the 20% would go higher. But what do I know?

Maybe that’s the answer for GREs. If you cannot build a business servicing a portfolio of countries with a combined GDP of at least 30% of global GDP, then do not do it.

The largest Abu Dhabi sovereign wealth fund, the Abu Dhabi Investment Authority (Adia), recently announced its results with the 20-year average return dropping to 6.5% from 7.4% and the 30-year average dropping to 7.5% from 8.4% the previous year. The drop was attributed to historical returns dropping off the rolling average, financial speak for saying the 1995 and 1985 returns were outsized ones that had pulled the average up and were now no longer being counted.

Let’s take a closer look at that. In the absence of detailed data, we have to make some assumptions which might not give the exact answer but should give the correct direction. The first is that it is commonly accepted best practice to use a benchmark for investment returns. A return of 40% might seem high, but it isn’t so great if the market as a whole grew by 60%. Similarly, a 2% return might not seem so good, but it is great if the market lost 20%.

One decent benchmark is the MSCI World Index (WI). Adia is clearly a global investor given its size and sophistication, so looking at a world index developed by the global leader in investment indexes makes sense. Last year’s return on the WI was minus 0.3%. We’ll call it 0% for our purposes.

If we assume, and this is a big assumption, that Adia matched the index performance last year, then mathematically the drop-off returns for 1995 and 1985 had to be 18% and 29%, respectively, to alter the rolling 20-year and 30-year returns to the extent they did. The 1995 WI return was 21.3% in 1995 and 41.7% in 1985. This supports Adia’s assertions and points to an outperformance last year.

But what about performance across the years? The WI has a 30-year return on investment of about 9.6% a year compared to Adia’s 7.5%. That means that US$100 billion invested with Adia would be about $775bn today. The same amount invested in the WI would be about $1.4 trillion today. That’s nearly double. On a 20-year basis the index again yields a better return, but not by as much.

Some might argue that the WI is the wrong index to use to benchmark Adia’s portfolio. But maybe it was the wrong portfolio to construct given the sizeable outperformance of the passive index. Post hoc ergo propter hoc, anyone?

The available data indicates that passive index investing might be the more effective strategy. But how does Adia stack up against other SWFs? It is hard to do an exact comparison, as different SWFs have different mandates. But if we look at Norway’s SWF, the largest in the world, it provides an 18-year history. If we use the WI to plug in the extra two years so that we can compare it to the 20-year Adia return we get an annual rate of return of 6.5%. So about the same.

As an aside, I saw the names of two UAE-based asset managers that Norway invested in – Ajeej Capital and Rasmala Asset Management. Well done.

The analysis of everything in this article, as the IMF likewise admits in its report, is less than perfect because of the understandable lack of full information. It is yet another reason for an effective governance body to oversee, coordinate and direct the individual and collective performance of the larger GREs and SWFs, as well as their effect on the domestic economy. After all, the boards of these entities overlap in many cases, so reducing their size to become effectively executive subcommittees and concentrating the main board members in the overall governance body could be more efficient, and it would make further rationalisation such as the Ipic-Mubadala merger much more effective.

I do believe someone once mentioned the idea of a Supreme Investment Council?

This article was originally published in The National.


  1. […] to the above two points is that the government or its related entities, the GREs discussed in my previous article, might be crowding out private lending. Government lending increased by 132 million, a 0.1% […]

  2. Great post – great analysis Dr Sabah. Thanks a lot.

  3. There could well be a variety of data-related shortfalls in the GRE vs Private sector ROA comparison. GRE’s are typically much larger businesses and so growth rates are typically lower. GRE’s also own assets that sometimes have more systemic/political/social importance vs a focus on high returns.
    We aren’t told the sector weightings of these GRE portfolios either. For example, the private sector is where most of the service economy exists. Service businesses typically have much higher ROAs than heavy-industry or capitally intensive businesses.
    These factors alone could explain a big component of the difference.

Comments are closed.