The Hunsader Follies

The Hunsader Follies

Since I originally identified problematic claims in Bugs, Features, and Aggressive Incompetence, the saga of updates surrounding Eric Hunsader’s comical attempts to hide his shoddy research and lie his way out of getting caught was becoming so long I moved it to this page where I can maintain the ongoing proofs of malice (in reverse chronological order) for posterity without cluttering up my original posts. Enjoy the lols.

Update 9/23/2014: Fwd: Fwd: Fwd: YOU’LL NEVER BELIEVE

Earlier this summer, Captain Nanex started to realize that he had cried wolf one too many times when financial journalists completely stopped taking his calls — or in his case the near-psychotic emails he would spam to unwitting reporters a half dozen times a day. I imagine this type of industry-wide ostracizing is mentally taxing for someone who takes everything personally. Indeed, with his so-called research now considered no more relevant than an episode of Ghost Hunters, he has turned to broader political punditry in order to maintain some relevance (and perhaps friends?).

The most recent epic blunder comes from re-posting a decade-old spam email:

nanex-congress2

Unfortunately for our well-coiffed companion, a quick google of any one of those lines immediately turns up results showing it is a complete hoax from 1999.

But like every Hunsader Folly™ this gets even more pathetic. Unlike our grandmothers who we understand couldn’t know any better when receiving a FWD: FWD: FWD: topic like this, Eric actually sought out and posted this nonsense as evidence to support yet another false claim he made just hours beforehand:

nanex-congress3

The tweet links to a personal webpage of a professor of sociology, David. J Hanson, who provides no citations to back up his purported legal expertise. Coincidentally, Hanson’s personal pile of unsupported schmuckiness has existed in the same format since the late 90s (perhaps a trend?).

But we have Google and Wikipedia. We can clearly see that the Speech or Debate Clause does not grant immunity to Congressmen for felonies: no further analysis necessary. Even the more curious lay-person like myself can go just a link or two below the first result and find that Hanson’s and Hunsader’s bastardization of the law is unquestionably without merit (emphasis mine):

The Supreme Court interpreted the language “in all Cases, except Treason, Felony, and Breach of the Peace” to encompass all crimes. Williamson v. United States(1908). Tracing the origins of the clause to parliamentary privilege, the Court found this identical language was used to qualify Parliament’s privilege from arrest so that the members of Parliament were not immune from criminal prosecution. The Court concluded that the Framers’ use of the identical phrase, without any explanation, indicated that Congress’s privilege was to have the same limitation regarding criminal actions as did the parliamentary privilege from which the language was borrowed. The clause, therefore, does not provide Congress with any immunity from criminal prosecution.

And so, this is where we find Dear Eric in the Fall of 2014: trying to support hoaxes with other hoaxes.

Keep chasing those ghosts, big guy.

Update 6/13/2014: Eric Learns New Word, Misuses It

This is cute. In a brand new and still completely wrong account of the Flash Crash, Hunsader accuses CFTC Chief Economist Andrei Kirilenko (now MIT professor) of “data mining” his analysis so that HFT firms don’t look so bad. Comically, Eric’s whole outrage could have been avoided had he been willing to read a single footnote. He posts:

Not convinced? Here’s how HFTs and Intermediaries were split in Andrei Kirilenko’s paper (CFTC Chief Economist and co-author of SEC/CFTC Report), which was published the very next day after the Flash Crash Report. From page 12:

From the CFTC Flash Crash Report

Kirilenko doesn’t use 3%, he uses 7%. And no surprise, the HFT group in Kirilenko’s paper isn’t nearly as angelic as the HFT group in the SEC/CFTC report.

But note how Kirilenko’s numbers don’t add up either: before dividing, there were 195 Intermediary accounts (16+179). 16 out of 195 is 8.2%, not 7%. Did Kirilenko also experiment with numbers to see what fit best?

This looks like a case of data mining to us.

Hunsader’s ineffable incompetence allows him to hold the belief that the CFTC’s Chief Economist doesn’t know the difference between 8% and 7%. Adorable. That same incompetence also prevents him from reading the footnote in his own picture. Here’s footnote 6:

2014-06-13 16_06_39-www.nanex.net_aqck2_4650_SSRN-id1686004

Perhaps Eric did read this footnote but simply didn’t understand what it meant. Allow me to explain. Classification in statistics often deals with how to place an individual into a single category based on many different variables — e.g. insurance companies classifying someone as “high risk” or “low risk”. There are many techniques for classification and some can be pretty math-fancy. Luckily this one is straightforward.

In the CFTC report, they developed a classification system to discern “High-Frequency Traders” from “Intermediaries” using data from May 3 – 5. Classifying the top 7% of intermediaries as HFT’s represented a statistically significant difference in trading activity between the two categories. It’s not the 7% itself that is the cutoff, it’s the trading activity (which just happened to be 7% for that day). In the CFTC’s classification system, that level of trading activity becomes the cutoff point for whether to classify someone on May 6 as an HFT or an Intermediary. May 3 – 5 was the calibration period.

From Table II in CFTC Paper

Adapted from Table II in Kirilenko Paper

You’ll see that 15 HFT traders during the calibration period accounted for around 3 times more trading volume than the 189 intermediaries. This is the significant difference in trading activity that created the cutoff point. When those same rules were applied to May 6, the relative ratios of the volume and trade statistics were about the same but the group sizes changed slightly.

This is actually the exact opposite of data mining. This is the correct way to use “in sample” versus “out of sample” data. Clearly Eric needs to understand what he’s looking at before crying foul. He might even be surprised to find that the official CFTC report features the exact same table. Nothing spooky going on here.

Update 5/12/2014: The Lies Continue

Recently Hunsader has caught a lot of criticism for his piss poor analysis of the Flash Crash. In particular, Nanex claims that the firm originally blamed for sparking the Flash Crash could not have done it because its trading algorithm could not have placed aggressive orders — despite the official CFTC report saying the exact opposite. So people with actual market data experience pointed out that this would be easy to check as the CME sends the aggressor tag in every traded outright contract. Unfortunately Nanex’s poor data doesn’t have this information. In self-defense, Hunsader told another bold-faced lie:

This is what a lie looks like

This is what a lie looks like

Basically he tried to claim that the ability to figure out aggressive orders wasn’t available at that time. Unfortunately for the big man, responsible Tweeters were all over it within minutes.

lol @ nanex

lol @ nanex

Since then Hunsader has been nervously deleting all of his tweets about this issue.

Update 5/09/2014: From Silly to Desperate

Sometimes you have to admire the stalwart ignorance of the loud and proud. Coming fresh off of his pants-down punishment for data fraud, Hunsader tries to chastise Dave Lauer for questioning Nanex’s most recent batch of phoney research. As if any sane person needed further proof that Eric Hunsader has absolutely no idea what he’s talking about, he makes this claim:

It’s not just bold. It’s not just stupid. It’s not just wrong. This is willful and malicious abuse of open, honest research and the people who attempt to conduct it. Nanex has shown itself to be far beyond just a simple group of clueless self-promoters.

Update 5/07/2014: Things Get Silly

Twitter is quite a remarkable tool. Not only was I able to publicly confront (and embarrass) Hunsader, but at the same time I was reaching out to a community of extremely knowledgeable and interested people. Nice to meet everyone!

Similar to when a child gets caught in a lie, Eric spent several weeks responding to my inquiries with dickish non sequiturs like this:

(hint: I never made that claim)

(hint: I never made that claim)

But soon enough, interested parties began investigating the fidelity of Nanex “research.” He finally retracted after numerous other people pointed out that he’s blatantly lying. He has removed the post from his site as of this update. Again like a child, he avoided me when admitting wrongdoing:

2014-05-08 22_20_02-TweetDeck

(deleted. not updated)

Although his retraction tweet is false in several ways, it’s over. I decided to close this case with panache:

2014-05-08 22_27_21-TweetDeck

(I take full responsibility for my immaturity)

Update 4/16/2014: Proof Nanex Is A Fraud

Given that we don’t have access to individual order data, I wondered what it would take to actually test Hunsader’s claims. Markets are generally slow at that time of night, so I figured I would concede if 5,000 contracts were cancelled in a single update. But what I found was even better.

Hunsader's Claim

Hunsader’s Claim

The problem: there wasn’t 5,000 contracts to cancel! This person had been getting filled on his order the entire time.

can't possibly cancel 5,000 contracts

can’t possibly cancel 5,000 contracts

I have attached the tick by tick order books for ESZ3 on Wed, 16 Oct 2013 07:00:00 to just after the cancellation: OrderBook ESZ3. Hunsader’s analysis is fraudulent. QED.

One Pingback/Trackback