One of the mantras my supervisors taught me on the trading desk was not to make other people smarter - if you hear someone say something that you know to be incorrect, assuming they don't work for your firm, you let them go on with their misconceptions. Knowledge is money.
I always had a slight problem with this, since I hate rampant stupidity, and now, since I'm no longer trying to profit off the ignorance of others, I am eager to shed some light on some misconceptions that are spreading amongst some blogs that I read daily, enjoy immensely, and find tremendously valuable.
Tyler Durden at ZeroHedge has quickly made a name for himself as an intelligent and detailed blogger in the financial realm. TD's deep investigative insights and tireless work ethic have earned him a rabid following for his blog, where he posts detailed, insightful posts several times a day. However, it's clear to me that TD is a fixed income guy at heart, and that his equity observations are frequently based on conspiracy theories or erroneous conclusions.
What's dangerous is that in this day and age people seem to believe everything they read on the internet. Financial bloggers have done a MASSIVE service in educating the public (or those who want to be educated at least) but that doesn't mean that everything that is written is the truth. Even more dangerous is when one blogger writes something erroneous, and others pick up on it - spreading the errors amongst their own rabid blog readership.
So, Tyler Durden at Zerohedge has been on Goldman Sachs's case for a while about their massive increase in principal program trading volume. The numbers are published publicly by the NYSE, based on a report the broker dealers file nightly called the DPTR (daily program trading report). Principal trading volume is volume that the firm trades for its own account. TD's thesis is that Goldman Sachs is manipulating the market via fancy computer trading models, and ripping off the public. Now, interestingly, Tyler's posts gained so much traction that
Goldman Sachs actually replied to them, explaining that the increase in volume was the result of GS participating in a new NYSE program called the SLP - supplemental liquidity program - where Goldman provides liquidity by posting bids and offers on the NYSE, and is compensated by the NYSE by receiving a tiny "rebate" whenever a counterparty accesses their quote.
To simplify, Goldman Sachs narrows the bid-ask spread by posting inside quotes, and they get paid a fraction of a penny if someone trades against their market. There is a temptation by the uninformed to conclude that sophisticated computers are controlling the market and ruining the world, but the truth is almost certainly that most investors are benefited by this added liquidity, as it lowers their cost of execution (in the form of a narrower bid-ask spread).
But that's not even what I wanted to write about today!
Today, Tyler Durden included the following quote
in his post about this week's NYSE Program Trading numbers:
"Zero Hedge is compiling materials to demonstrate the phenomenal gamble CS is taking by being the largest holder of the ETF-underlying pair trade. The ensuing implosion, once the market loses the invisible futures bid, will likely destroy Switzerland's second biggest bank and likely take down the country with it.
Probably most notable is the screaming increase in overall program trading, from 30.7% of all NYSE volume to 40.4%! Virtually every broker saw their Principal PT operations double week over week: seems like everyone is brokering those ETF trades now. Poor SPY and IWM are being mangled 10 ways from Sunday nowadays."
Never mind the massive increase in the weekly principal transaction numbers - it's due to June's quarterly expiration, where brokers trade massive baskets of stock against expiring futures and options. I can tell you already that next week's program trading statistics will show a large increase in the "customer facilitation" numbers, because that's how the trades for today's annual rebalancing of the Russell indices will be coded. What I want to talk about is the misunderstanding about the "ETF-underlying pair trade." First, let me take a quick tangent.
Another blogger who I enjoy for his relentless attempt to try to expose the wrongdoings of the authorities is
Karl Denninger. Denninger had some additional insights related to
my recent post on AIG dumping its assets on the Fed,
pointing out that the Fed's actions here are outright illegal. However, Denninger
also jumped all over the ZeroHedge quote about the ETF-underlying trade and its relationship to increased program trading volume:
"For those who aren't savvy in this stuff what's going on here is a pair trade between the underlying instrument(s) and the ETFs on the exchanges. This is an arb play and it works until it doesn't - for an example of "doesn't" in the single-name world witness what happened to VW/Porsche earlier this year when the arb speculators on their merger got rammed, or those hedgies who were playing the Citibank preferred-conversion arb earlier this year.
These are allegedly "hedged" transactions in that there is an alleged unbreakable correlation that protects the person doing it from loss. In truth there is no such thing as an unbreakable correlation and the alleged "protection" against getting reamed is illusory. This is the same sort of "genius trade" that was run with AIG's CDS positions - remember the claim that "we're unlikely to ever see a loss"? "
So, now it's time for a lesson about arbitrage. There are many types of arbitrage. The simplest, which I still remember reading about in the famous "Gold Book" that UBS Warburg provided all of its new employees back in the 1990's as an introduction to options and markets theory, was "if you can buy gold for $400/oz in New York, and sell it for $450/oz in London, that's called an "arbitrage." This is a pure arbitrage. You buy gold for $400 and sell it for $450. You profit $50, minus the cost of transporting the gold.
Another type of arbitrage is merger arbitrage: Let's say company ABC is buying company XYZ, and that ABC is offering shareholders of XYZ 1 share of ABC stock for each share of XYZ they currently own. The prices of the two stocks should converge as the deal nears completion. There will be a spread between the two stock prices, depending on the risk of the deal (and some other factors like the dividends that will be paid out until the deal closes, and some financing costs). Merger arbitrage has risk - the deal might fall apart - if you buy XYZ shares and short ABC shares at a higher price, you are very much NOT guaranteed to capture that spread. The Citi vs Citi Preferred trade that Denninger mentioned falls into this category - the traders got hurt because the spread widened, but if the conversion actually goes through, they will recapture the spread they have lost (although there are still issues with this trade due to the high costs associated with borrowing C stock to short it for a longer period of time than initially expected.)
Then there are convergence trades like the ones Long Term Capital Management made famous, based upon traditional trading relationships between two assets. LTCM would buy the "cheap" asset and sell the "expensive" asset and expect the two to converge. Clearly, there is no assurance that the prices will converge here - and when you add massive leverage to the mix like LTCM did, small adverse price movements can have disastrous consequences. The Porsche-VW stub trade which denninger mentioned is most related to this class of convergence trades.
We also have share class trades - for stocks like Berkshire Hathaway, which have two classes of shares: BRK/A and BRK/B. In Berkshire, the B shares are supposed to trade, I think, at 1/30th the price of the A shares - but there is no mechanism for which you are entitled to exchange shares between the two share classes. Thus, if you put on a trade seeking to profit from the disparity in value between the A and B shares, you have to wait and hope the prices converge so you can reap your profit.
This brings us to the ETF-Underlying pairs trade. Now, forgive me if I'm putting words in Tyler Durden's or Karl Denninger's mouths, but it seems that they are referring to the trade where a broker, say CS, buys an ETF (like IWM) and shorts the underlying basket of stocks against it. The IWM is composed of the 2000 stocks in the Russell 2000 index, and the trust that issues the IWM owns these underlying stocks against the IWM shares it has issued. Lately, CS has shown up as a large holder of IWM and SPY, as well as some other ETF's.
Today, both ZeroHedge, and, jumping on the bandwagon, Karl Denninger, decry this as a potent of doom - a potentially lethal arbitrage that could blow up in CS's face. There's only one problem - if CS is buying the ETF's and shorting the underlying baskets of stocks, as expected, there is no risk. This trade is not like the trades I mentioned above for one specific reason - ETF's can be created and redeemed daily: If you own a chunk of IWM (the minimum unit is 50k shares) you can take your IWM shares and deliver them in to the trust (aka: redeem) in exchange for the corresponding number of shares in each of the 2000 underlying stocks (which you then use to cover your short positions). Similarly, if you are long the underlying components, you can deliver them in to the trust (aka create) and they will give you IWMs. It's precisely this fungibility that completely nullifies TD's and Denninger's fears. It's a very simple concept, and it's not at all like the convergence trades above, where you cannot control the collapsing of the two legs of the trade.
Denninger isolates all of the comments on his posts in his forums, and regarding this topic, he elaborated:
"Let's say that there's a "divergence" of a nickel betwene the two instruments. You short one and buy the other, allegedly pocketing the nickel. The theory is that the correlation will remove the spread, at which point you close both positions."
Again - with ETF's you can create and redeem at will - so you don't have to rely on any "theory" or "correlation." You can make it happen yourself.
I'm somewhat surprised that two bloggers as intelligent as Tyler Durden and Karl Denninger erred in their assault on this concept, but I was even more troubled by the comments in each of their respective posts, which show that the misinformation spreads rapidly, as their readership takes what they write as truth. In the ZeroHedge post, I offered a possible explanation of why CS has these trades on: "CS probably has some sort of funding efficiency where they lend out the long ETF and charge a borrow rate that exceeds their cost of capital (courtesy of near zero short term rates from the Fed)... if this funding efficiency goes away they just redeem the ETF to cover their short stock position."
Financial blogging has been a great breakthrough in the last several years, with a plethora of blogs offering an educated, insightful look at issues that the mainstream media could never understand. Many bloggers are actually former professionals in the topics that they blog about, which gives them unique insight into topics that journalists could not attain. Still, the flaw of the internet is that it spreads all information equally, and as a reader it's up to you to verify the thought processes and make sure you understand them before you take them as fact.
-KD
note: I have never worked at GS or CS. I have no incentive to defend GS - I don't particularly like the "walk on water" status they have attained. Also, I do not intend for this post to be an assault on TD or Denninger - I think they both write excellent blogs, they are just wrong on this topic.