It seems that we have still a long ways to go to settle on a definition, but Stephen Jen from Morgan Stanley has a good attempt located here:

Summary and conclusions

There is no universally agreed definition of sovereign wealth funds (SWFs).  In this note, we try to better define what SWFs are. 

Characteristics of SWFs

Mr Clay Lowery – then Acting Under Secretary for International Affairs – gave a speech on SWFs on June 21, 2007, in which he presented his definition: A SWF is “a government investment vehicle which is funded by foreign exchange assets, and which manages these assets separately from official reserves”.  This definition is close to what I have in mind.  To me, a SWF needs to have five ingredients:

1.    Sovereign;

2.    High foreign currency exposure;

3.    No explicit liabilities;

4.    High risk tolerance;

5.    Long investment horizon.

There are close cousins of SWFs.  Official reserves are related to SWFs, as are sovereign pension funds (SPFs).  These three categories of public funds have different characteristics, but are not necessarily mutually exclusive.  Rather than providing a more precise definition of SWFs, I believe that it would be more accurate to describe what these funds are in a diagram. 

Official foreign reserves are, by definition, 100% in foreign currencies.  They have no liabilities explicitly attached to them, though, indirectly, they are financed by domestic government bonds used to finance the foreign exchange interventions in the first place.  

In my definition, SWFs don’t need to be 100% in foreign currencies, but should be mostly in foreign currency terms.  For example, Singapore’s Temasek Holdings, Malaysia’s Khazanah Nasional BHD and Canada’s Fond des generations (Quebec) are not 100% held in foreign currency assets, but we still consider them SWFs, as they have high exposure to foreign currencies. 

Sovereign pension funds, on the other hand, could have an even lower foreign currency content.  Japan’s GPIF and the US Social Security Trust Funds, for example, have 0% and 13%, respectively, of foreign currency content. 

Also, SPFs tend to have either explicit or implicit pension liabilities attached to them, while SWFs do not.   

There could of course be some overlap between SWFs and SPFs.  The best example is Singapore’s GIC.  But even Norway’s GPF could be another example.  

Further, SWFs and their cousins span a spectrum along two other dimensions: their risk tolerance and the length of the investment horizon. 

Due to their need to have liquidity and security, official foreign reserves have very short investment horizons (which is different from the concept of ‘duration’) and low tolerance for credit risk.  This is why most of the official reserves are held in sovereign bonds in large and liquid markets.   

Hedge funds, while they may also have a short investment horizon, tend to have very high tolerance for risk.      

Both SWFs and SPFs tend to have a longer investment horizon, which may be decades long.  SWFs, as I argued in a previous note, should have a higher risk tolerance than SPFs. 

My thoughts

Rather than being well-defined and distinct from other types of funds, there is a great deal of overlap between SWFs and their close cousins.  Rather than coming up with a complicated definition of a SWF, I think it is more useful to highlight their distinguishing traits, as compared to the other types of funds. 

I have these thoughts: 

Thought 1.  Regulatory discrimination unjustified.  There has been much political angst about SWFs.  However, seen from the perspectives described above, it does not seem to make sense for regulatory authorities and politicians to single out SWFs.  SPFs, private pension funds and hedge funds may also have issues regarding transparency.  Further, SWFs could easily invest in hedge funds and private equity funds to circumvent restrictions.  It appears that a comprehensive approach towards regulating the whole array of funds may make more sense than singling out the SWFs.  For example, there are many examples of SPFs holding controlling stakes in major airports and infrastructure projects.  Why then should anyone be worried about SWFs doing the same?

Thought 2.  The purpose of analyzing SWFs.  Since these different types of funds overlap each other, how one defines SWFs depends to some extent on the purpose one has in analysing these funds.  For example, I am mostly interested in the impact of SWFs on risky assets, not on the regulatory or policy implications.  Including Singapore’s GIC and Malaysia’s Khazanah Nasional BHD in the SWF table or the SPF table really does not make a big difference.  Similarly, Australia’s Future Fund and New Zealand’s Superannuation Fund could be in the SPF or the SWF table, because how these funds are categorised does not alter their impact on the risky assets. 

Thought 3.  SWFs a source of market stability.  There is a presumption that SWFs could become a source of instability, disrupting and crowding out private capital flows.  I disagree.    We believe that   SWFs could have a much longer investment horizon and a higher tolerance for swings in P/L.  When the global equity markets corrected in 2000, Norges Bank was a heavy buyer of global equities.  I would not be surprised if the SWFs already played a meaningful role in September, facilitating the recovery in EM equities and equities in general.  Having such a different ‘temperament’ from private funds, SWFs should reduce the risk of ‘herd behaviour’. 

Also, a notion we have been stressing for the past year is  that a conversion from official reserves into a SWF would significantly raise risk-taking and yield a bias favourable to equities at the expense of bonds. 

Furthermore, market efficiency is a function of market liquidity.  To the extent that SWFs improve market liquidity, particularly in a way that is not ‘herdish’ like other types of short-term capital flows, SWFs should be a positive factor for markets in general.  Regulators and politicians should recognise these positive characteristics of SWFs and weigh them against the concerns about governance, transparency and natural security. 

Bottom line
In this note, I elaborate on my definition of a SWF.  In my view, a SWF needs to have five characteristics: (1) sovereign; (2) high foreign currency exposure; (3) no explicit liabilities; (4) high risk tolerance; and (5) long investment horizon.