Apr

26

This price movement of June bond futures on the morning of Apr. 23 taught me a lesson.

From 7:52 AM to 8:55 AM (slightly over an hour) the market went from 111-10 to 111-01. During this period it traded 19,459 contracts. There were no announcements during this time frame.

Then from 8:56 AM to 9:57 AM the market completely reversed its direction and went from 111-10 to 111-01, exactly where it began two hours previously. The additional contracts traded in that time frame: 20,469. Almost the exact amount as on the way down.

Why would market participants all of a sudden change sentiment, when there were no announcements? What makes participant bias change so abruptly without news?

Robert Ray replies:

A nine tick Lobagola? Take that same move from the perspective of someone that wasn't watching every tick and it would appear that not much at all went on as the price was the same two hours later. There is a meal here in how one perceives things.

Edward Talisse remarks:

This behavior happens all the time, not only in US but in Bunds and JGBs. It's the hedging of new issue deal flow. As corporate bonds are priced, dealers (read: swap desks) trade the order flow but usually end up flat. It has nothing to do with availability of new information.

George Zachar adds:

Deal flow is information, and gaming the hedges and their lifting is a major part of the debt market's micro-process.

Phil McDonnell explains:

A price quote is for a completed transaction. It is always between a buyer and a seller. So the number of buyers always equals the number of sellers — no exceptions. Only the price adjusts. So logically the number of long contracts equals the number of shorts always, all futures markets — no exceptions.

You can infer something about the initiator of the trade. He is often a market order coming from off floor. The market order will cross on the floor (or in a computer) with the current bid or ask. So a down tick usually means an off floor trade crossed with the bid. An uptick often indicates an off floor market order crossed with the current best ask. This is only the commonest case and must be tempered with the realization that limit orders will appear as bid/ask quotes as well and may be confused with market maker activity. Also you cannot know if open interest is increased or decreased by a single trade, but you can track it on a net basis over longer time periods.

Victor told the tale of the elephants always returning by the same path in his book EdSpec. It was a story originally told by Lobagola. The story holds true for markets as well. Markets tend to retrace the same ground — often many times.

Is there statistical evidence for this? One need look no farther than Doc Castaldo's recent post on the Pythagorean scale and markets. His data showed that markets exhibit small changes far too often for it to be chance. This is the salient feature of speculative markets. It happens all the time. Huge amounts of money are made and lost on the very numerous small change days.

Consider the idealized model of a market with a single market maker. He quotes 100 to 101. Someone sells to him at 100. So he drops his quote to 99 to 100. The very act of dropping the quote inspires more selling. He drops his bids to 98, 97, 96 then 95 in succession as more sales come in. His average cost is about 97.50. Now at 95 a funny thing happens. He hits somebody's threshold of pain, whose stop is executed at 95, and our market maker winds up too long. Now he holds the price firm or even raises it.

On the perception of firming or even rising prices speculators start to nibble. Our market maker now slowly raises prices back up to where they were before. Only this time he is supplying at the ask price. So he makes his spread which he tries to maintain at one point. By the end of the day the quote returns to where it was before. Our market maker has sold his inventory at an average of 98.50. The market has done a Lobagola down then up. The news reports the market was unchanged today and everyone yawns. But our intrepid market maker made his spread going down and then back up. He can afford to eat at Delmonico's yet another night.

Sam Marx adds:

As a former market maker on the floor, I can say that this description is a good approximation of what happens. That's why the distribution of prices is higher at the middle than the normal distribution. The market maker is more confident within the existing range.

Also, when there is an large influx of sell orders, the market maker steps aside, buying smaller quantities, a minimum number of lots at each lower price to perform his function, and lets the price really drop. When buy orders start coming in, or when the sell orders stop, he starts buying. That's why the price distribution is lower than the theoretical normal distribution a short distance from the middle of the curve. A leptokurtic distribution.

J.T. Holley extends:

I'm looking at long bonds today. The UK 50 year yields 4.1% and the French Euro 50 year around 4.0% Does this mean that the US long bond is going to 4%? Has anyone with a scientific bent studied/counted the ratios/differences of these instruments' yields?

Faisal Essa responds:

The UK and EUR long bonds are said to trade at those levels because of the local pension and insurance company law changes that have forced pension funds to match their long duration liabilities with long duration, high quality bonds. This has led to reduction in allocation to equities and a shortage of long bond supply relative to demand. To make matters worse, restrictions on currency exposure and derivatives overlays force the funds to stay in their own market rather than buying other countries' long bonds. This legal framework is quite unfortunate for Dimsonian pensioners.

This situation (along with changes in US pension fund law) does have some influence on US long bonds and long TIPS.

Charles Sorkin suggests:

If the media decide to exploit the notion that the American homeowner needs to be bailed out, bonds could fly. An interesting hedge (although extremely difficult to model and get the ratio correct) would be a short bond position hedged with long support-class POs. Difficult for the small investor to find, but some pieces have been floating around lately in the upper $30s to upper $40s on long paper. Such securities could return your principal within two years on a bond rally of 100-200 bps. 

Archives

Resources & Links

Search