Before the Joint CFTC-SEC Advisory Committee on Emerging Regulatory Issues

 
 
 
 
  
    
  
 
 
  
Statement of
 
Michael A. Mendelson, AQR Capital Management LLC 

Before the Joint CFTC-SEC Advisory Committee on Emerging Regulatory Issues 

August 11, 2010 

Chairman Schapiro, Chairman Gensler, and members of the Joint Committee, my name is 
Michael  Mendelson.    I  am  a  Principal  at  AQR  Capital  Management,  an  investment  
management  firm  that  for  the  past  twelve  years  has  managed  assets  for  pension  funds,  
endowments, and foundations and now also manages public mutual funds. 
Thank you for inviting me today to discuss our experience of the events of May 6.  The 
“Flash  Crash”  highlights  a  risk  in  an  otherwise  well-functioning  US  equity  market.  
While AQR and the great majority of other investors managed to avoid damage from this 
event,  unfortunately  not  everyone  did.   We  can  reduce  the  likelihood  of  a  repeat  Flash  
Crash and the work of this Committee and the extensive efforts of the staffs of the SEC 
and CFTC may be the most important steps in that effort.  
AQR employs quantitative methods in most of its investment strategies.  We invest in a 
wide  variety  of  instruments,  including  US  equities.    Our  holding  periods  are  typically  
months  to  years.   Some  of  our  investment  strategies turn over every few days, but none 
would be considered “high frequency”.  We are liquidity seekers – though we don’t use 
market  orders  -  and  rely  on  liquidity  providers  to  perform  their  essential  function.    In  
many  of  the  markets  in  which  we  trade,  such  as  fixed  income,  liquidity  is  provided  by  
dealer  firms  whose  ability  to  provide  liquidity  rises  and  falls  with  the  health  of  the  
financial system.  But in the US equities market, liquidity is provided by a broad base of 
participants.  This was a great benefit to all investors during the most difficult weeks of 

  
   
 
  
the  financial  crisis.    Our  exchange-traded  markets  performed  admirably  and  those  
responsible for that, from the small electronic market makers to the regulators who led us 
to  a  competitive,  broadly  democratized  market  structure,  should  be  proud  of  this  
achievement. 
At  AQR,  we  build  safeguards  into  our  trading  processes  and  have  human  oversight  of  
them,  important  steps  for  protecting  client  assets.    On  May  6,  our  trading  staff  noticed  
early on that the market was potentially disrupted and shut down our equity trading.  We 
suffered no “busted” trades, we avoided trading at dislocated prices, and we were able to 
complete   the   overwhelming   share   of   portfolio   transactions   planned   for   that   day.   
Nevertheless, May 6 highlights risks in the trading ecosystem that need to be managed. 
I would like to highlight three issues. 
First,  questions  remain  about  the  cause  of  the  Flash  Crash,  but  we  know  there  was  
significant   negative   macroeconomic   news,   very   heavy   trading   volume,   substantial   
liquidity  demand  from  market  sellers,  trade  reports  that  appeared  to  be  erroneous  or  
delayed,  and  a  de-linking  of  our  trading  centers.    With  liquidity  providers  experiencing  
large P&L moves while fearing they were flying blind without reliable market data, it is 
easy  to  understand  why  they  would  have  felt  compelled  to  withdraw  their  limit  orders.  
Meanwhile,  liquidity  demanders  continued  to  send  market  orders,  unaware  that  the  
typically deep limit order book, wasn’t.  I want to emphasize the importance of liquidity 
demanding investors being “unaware” of the disappearing liquidity. 
Second, had some of the weak links been stronger, what alternative course could events 
have  taken  that  day?    Perhaps  the  evaporation  of  the  limit  order  book  was  actually  our  

  
  
 
  
 
good  fortune,  as  the  subsequent  shocking  trade  reports  screamed  out  to  market  sellers  
“stop!”    Without  that  loud  blast,  selling  may  have  continued  unabated,  causing  a  real  
crash  from  which  it  would  have  taken  far  longer  than  15  minutes  to  recover.    The 
effective  clearing  of  the  limit  order  book  might  have  acted  much  like  a  circuit  breaker,  
albeit  a  very  sloppy  one.  Better  market  data,  better  exchange  coordination,  and  
additional rules might have prevented the flash crash, but might have enabled a real one. 
I don’t know.  Careful analysis of market data may yield an answer, and I encourage the 
Committee  to  work  with  industry  participants  to  explore  this.    We  need  to  understand  
what  role  demanders  of  liquidity  had  on  May  6  and  perhaps  consider  steps  to  better  
inform those participants of the live, aggregate supply of liquidity.  
Third, the complexities of our trading environment should give us pause and at least drive 
us toward seeking lightweight and simple solutions.  Toward that end, the current circuit 
breaker  pilot  program  may  be  a  good  start,  but  modifications  may  be  needed.    We  have  
seen  as  recently  as  last  Thursday  that  erroneous  trade  reports  can  halt  a  stock.    With  
broad access to the Trade Reporting Facility, there is too much potential for abuse, even 
catastrophic abuse.  So perhaps consideration should be given to a limit offer rule. 
Another  proposed  solution  is  to  impose  market-making  obligations.    This  will  increase  
costs  for  retail  and  institutional  investors  every  second  of  every  normal  trading  day  by  
reducing  the  availability  of  liquidity  providing  capital  and  increasing  its  risk.    Adding  
insult  to  injury,  on  those  rare  occasions  when  markets  are  severely  disrupted,  market-
maker obligations will accomplish nothing.  After all, the function of a market maker is 
not  to  buy  stock  at  the  wrong  price  as  a  market  is  crashing.    Market  making,  whether  

 
 
complete with a strong set of obligations or not, has never worked that way, and it never 
will. 
Market maker obligations come with special privileges and some markets may need this 
to encourage liquidity providers in the ordinary course of business.  But instead, here the 
suggestion  is  that  these  privileges  will  encourage  liquidity  provision  in  extraordinary  
times.  They won’t. 
Thank you 

Statement of
 
Michael A. Mendelson, AQR Capital Management LLC 


Before the Joint CFTC-SEC Advisory Committee on Emerging Regulatory Issues 

August 11, 2010 


Chairman Schapiro, Chairman Gensler, and members of the Joint Committee, my name is 

Michael Mendelson. I am a Principal at AQR Capital Management, an investment 

management firm that for the past twelve years has managed assets for pension funds, 

endowments, and foundations and now also manages public mutual funds. 

Thank you for inviting me today to discuss our experience of the events of May 6.  The 

“Flash Crash” highlights a risk in an otherwise well-functioning US equity market. 

While AQR and the great majority of other investors managed to avoid damage from this 

event, unfortunately not everyone did. We can reduce the likelihood of a repeat Flash 

Crash and the work of this Committee and the extensive efforts of the staffs of the SEC 

and CFTC may be the most important steps in that effort.  

AQR employs quantitative methods in most of its investment strategies.  We invest in a 

wide variety of instruments, including US equities.  Our holding periods are typically 

months to years. Some of our investment strategies turn over every few days, but none 

would be considered “high frequency”. We are liquidity seekers – though we don’t use 

market orders - and rely on liquidity providers to perform their essential function.  In 

many of the markets in which we trade, such as fixed income, liquidity is provided by 

dealer firms whose ability to provide liquidity rises and falls with the health of the 

financial system.  But in the US equities market, liquidity is provided by a broad base of 

participants.  This was a great benefit to all investors during the most difficult weeks of 



 

 

 

 

the financial crisis.  Our exchange-traded markets performed admirably and those 

responsible for that, from the small electronic market makers to the regulators who led us 

to a competitive, broadly democratized market structure, should be proud of this 

achievement. 

At AQR, we build safeguards into our trading processes and have human oversight of 

them, important steps for protecting client assets.  On May 6, our trading staff noticed 

early on that the market was potentially disrupted and shut down our equity trading.  We 

suffered no “busted” trades, we avoided trading at dislocated prices, and we were able to 

complete the overwhelming share of portfolio transactions planned for that day. 

Nevertheless, May 6 highlights risks in the trading ecosystem that need to be managed. 

I would like to highlight three issues. 

First, questions remain about the cause of the Flash Crash, but we know there was 

significant negative macroeconomic news, very heavy trading volume, substantial 

liquidity demand from market sellers, trade reports that appeared to be erroneous or 

delayed, and a de-linking of our trading centers.  With liquidity providers experiencing 

large P&L moves while fearing they were flying blind without reliable market data, it is 

easy to understand why they would have felt compelled to withdraw their limit orders. 

Meanwhile, liquidity demanders continued to send market orders, unaware that the 

typically deep limit order book, wasn’t.  I want to emphasize the importance of liquidity 

demanding investors being “unaware” of the disappearing liquidity. 

Second, had some of the weak links been stronger, what alternative course could events 

have taken that day?  Perhaps the evaporation of the limit order book was actually our 



 

 

 

 

 

good fortune, as the subsequent shocking trade reports screamed out to market sellers 

“stop!” Without that loud blast, selling may have continued unabated, causing a real 

crash from which it would have taken far longer than 15 minutes to recover.  The 

effective clearing of the limit order book might have acted much like a circuit breaker, 

albeit a very sloppy one. Better market data, better exchange coordination, and 

additional rules might have prevented the flash crash, but might have enabled a real one. 

I don’t know. Careful analysis of market data may yield an answer, and I encourage the 

Committee to work with industry participants to explore this.  We need to understand 

what role demanders of liquidity had on May 6 and perhaps consider steps to better 

inform those participants of the live, aggregate supply of liquidity.  

Third, the complexities of our trading environment should give us pause and at least drive 

us toward seeking lightweight and simple solutions.  Toward that end, the current circuit 

breaker pilot program may be a good start, but modifications may be needed.  We have 

seen as recently as last Thursday that erroneous trade reports can halt a stock.  With 

broad access to the Trade Reporting Facility, there is too much potential for abuse, even 

catastrophic abuse.  So perhaps consideration should be given to a limit offer rule. 

Another proposed solution is to impose market-making obligations.  This will increase 

costs for retail and institutional investors every second of every normal trading day by 

reducing the availability of liquidity providing capital and increasing its risk.  Adding 

insult to injury, on those rare occasions when markets are severely disrupted, market-

maker obligations will accomplish nothing.  After all, the function of a market maker is 

not to buy stock at the wrong price as a market is crashing.  Market making, whether 



 

 

complete with a strong set of obligations or not, has never worked that way, and it never 

will. 

Market maker obligations come with special privileges and some markets may need this 

to encourage liquidity providers in the ordinary course of business.  But instead, here the 

suggestion is that these privileges will encourage liquidity provision in extraordinary 

times.  They won’t. 

Thank you