This page has been archived and commenting is disabled.
One-In-A-Billion "Hiccups" Are Happening All The Time, Citi Warns Something Is Wrong
Earlier this year JP Morgan’s letter to shareholders, Jamie Dimon let it slip that there are some very disturbing things going on in today’s capital markets. Prices can gap in illiquid markets, Dimon explained, and that has the very real potential to spark a panic, causing illiquidity to spread to previously liquid markets. Dimon warned that one should not be fooled by relatively tight bid-asks; it’s market depth tells the true story, and as JP Morgan’s Nikolaos Panigirtzoglou will tell you, some markets (the Treasury market for instance) are getting quite thin indeed.
The dangers associated with a widespread lack of market depth are of course exacerbated by the presence of HFTs. This was on full display during last October’s Treasury flash crash. Here’s what Dimon had to say on the subject: "..then on one day, October 15, 2014, Treasury securities moved 40 basis points, statistically 7 to 8 standard deviations– an unprecedented move – an event that is supposed to happen only once in every 3 billion years or so." "Some currencies recently have had similar large moves," Dimon added, referencing the carnage that accompanied the SNB’s abandonment of the euro peg in January (as well as countless other flash crashes and rips) and presaging precisely what we’ve seen this week on the heels of China’s move to devalue the yuan.
The takeaway from all of this is not, as Dimon concluded, that statistics can’t be trusted, but rather that when things that are supposed to happen once every 3 billion years start happening once every three months, or every three weeks, then something is definitively broken.
Here, courtesy of Citi’s Matt King, is a look at some of the major one in a billion year events that have taken place over the last four years:
As King notes, either there’s a "widespread case of the hiccups", or something else is very, very wrong.
»
- 158836 reads
- Printer-friendly version
- Send to friend
- advertisements -
Math is imperfect, it is nothing more than man's feeble attempt to bridge his misunderstanding of natural law, which is indeed perfect. Natural law always asserts itself... always has, always will...and Natural law could care less about your PhD.
Well put!
Very well put.
Even if it's an MIT or Harvard PhD?
Especially if it's an MIT or Harvard PhD?
BS = bull shit. MS = more shit. PhD= piled higher & deeper.
Perfect? The entire Idea of Perfection is a human invention. Nature operates on the principles of Chaos, Complexity, and Perversity. And each in its own way is the antithesis of "Law".
The only real story the street cares about is the one about Sams dog dying. Forget the untold millions going to hell in handbasket, my fucking labrador has a wheezy chest and dont fetch no more.
When they started determining the economy on markets in stead of on budgets is when the world started to go to hell in a handbag.
[Example] We had a budget of $100 and made $200. So the economy is good and we have money to put aside.
[Not] Joe sold 10 bags of apples last week but sold 12 bags of apples this week, so the economy is good and we need to spend joes extra money right away on more apple trees. Whoops! Joe sold less apples this week so we need to dig up some trees. Whoops! Looks like joe is going to sell more apples next week so we need to plant more trees. Whoops! Joe didnt sell as manny apples as we thought but he almost sold as much as we thought so, just dig up a couple trees and leave some. Whoops! Holy Shit! Joe lost 1/2 his trees in a fire. I lost a shitload of money in Joes apples.
Get my drift? It aint working. The old man bankster idea is failing. Fast.
Make a budget, meet the budget. Growth leads to Hell and Hell alone.
But the whole point about apple trees is that they DO grow . . . slowly. You can't make decisions about planting trees (or investing in nations) based on monthly or annual conditions. You have to look forward and back for decades, maybe centuries.
The problem has to do with short-term analysis. NOTHING in the modern world is based on an adequate time-sample. Even "markets" like stocks, bonds, and gold, rarely consider fluctuations over more than a decade.
Humans live six to ten decades. Civilizations rise and fall in four cycles of roughly 50 years each. You CANNOT make adequate decisions based on the quarterly earnings or projections.
Bingo!!! And I would add that short term planning always leads to poor decisions for long term
So the first question that comes to mind is....
Before I ask it, let me recap. We had an event that is only supposed to happen once in a billion years. Therefore it could have happened 14 times since the big bang. The questions are:
"When did it last happen? Has it been less than a billion years; or are we overdue for a once in a billion year event?"
If it's been a billion years since it last happened, maybe it was time, so quit your griping.
If it's never happened then I say, "well it's about god damn time."
They are all searching, searching, searching....
Somewhere out there, a safe haven for their money must exist-- one which offers both security and ridiculul large returns.
Or so they believe, and so the search goes on.
The hiccups will keep right on happening, because theses events aren't outliers
anymore. History doesn't apply now because "we" are making up "our" own
reality as we go.
Just heard a Jesuit say tonight that we need firm banks.
The only solid bank they sell is death.
Finally something we can count on. Thanks for the reminder!
This "once in 3 billion years" stuff is bogus. Markets do not follow a Gaussian distribution, although they may mostly approximate it during "normal times". Therefore, you can say nothing about the probability of any move of any size.
Obviously the wrong distribution is being used, Jamie. These are 7-year events, not 3 billion.
Only complete idiots and servants of the sharks are still in these "markets".
mock-its.
What they are really afraid of is paper Ag and Au breaking. That will be a very very large natural number STDDEV event.
something is wrong alright.
It's the result of Banks counterfeiting hundreds of Trillions of Dollars and trying to pass it off as Money.
In Islam captured infidels are property. Property can not be raped. The story is a lie.
Jamie "let it slip" .... in writing no less. BS, the only thing that guy slips is ..............insert (haha) favorite Jamie dick joke
There is only one deviation responsible for this chaos - a group of deviants called the Fed!
those sigma are all meaningless, since they are sampling the dependent variables, which is the result of independent variables. citi needs to explore what the independent variables are. to sample the interest rates and money supply could be more useful.
it might have something to do with the number of participants(very few) and the lack of real sentiment having been replaced by computer driven algos. computers gaming computers.
And you just know that JPM goes hunting for those opportunities every day.
In other words, “We’re not banging the close, those moves are a natural result of market illiquidity.”
It’s getting harder and harder to separate explanations from excuses.
"when things that are supposed to happen once every 3 billion years start happening once every three months, or every three weeks, then something is definitively broken."
No doubt some things are broken, but probably more are not recognized or fully understood. The mathematical models don't (yet) contemplate as yet unknown and misunderstood factors. Don't put too much faith in models.
Dear Jamie, It's called "tail risk" you putz, and (as you should know but apparently don't) the bell curve statistical model never reaches zero no matter how many STD DEV you go out on the curve. Having said that, though, I hardly see the relevance of a 3 billion year event when all of these markets are only 40-50 years old; how do you know if what you've seen in your lifetime is truly the "norm"; maybe the norm is extremely volatile markets. I fail to see how a 50 year sample can equate to 3-4 billion year data period [which of course doesn't exist]. Your real bitch should be with the FED.
www.traderzoo.mobi
All of these horrendous events and still no meltdown. Seems like it demonstrates the resilience of the new world market and the central bank manipulations, not impending doom. BTFD.
One way HFT hurts is by creating the ILLUSION of liquidity.
If you know it isn't there, you can adjust, even if the adjustment is to not play the game, or in different size.
If you incorrectly think it is there, bad decisions can be made easily.
Discussing 'standard deviations' in market pricing depends on a couple of assumptions about the behaviour of markets. The first assumption is that the bid-ask are the result of independent actors approximating a consensus price. The second assumption is that the market behaviours can be compared over time. When there are 'low-probability' events occurring very frequently, it means that one of those two assumptions are violated. Either independent price estimation is broken - as, for instance, through market manipulation; or historic price movements are no longer comparable - fundamental shifts in market operation changing the price dynamic.
These frequent '7 and 8 sigma' events are a signal that the markets are behaving differently. It means previous experience in the market, previous expectations, are a poor guide to current expectations.