Friday, July 19, 2013

Why Do Treasury Traders Complain About Low Trading Volume??

On the one hand, for just about everybody trading in the UST market, there is always enough liquidity to put on or take off a position, in full, at any given moment.  How many traders trade more than 1000-2000 10yr contracts?  You can always buy or sell 1000 10yr contracts without moving the market.   While this is a problem for the primary dealers (who often bid / offer 1bln 10yr notes for their customers...the equiv of 10,000 10yr contracts), this is not a problem for the average day trader.

However, the price action patterns that occur when 50mm/bp (60 bln 10yr note equivs) trade in a 15minute time bar...vs when 5mm/bp trades in a 15minute bar...well, they are just different.  I know that statement is a little vague, but its difficult to put into words what we traders have learned via pattern recognition.   Everybody sees this activity in the screens, and the existence of this activity means more traders (large traders) are participating in the market.   Even though the UST market is the "deepest and most liquid bond market in the world"....there are still quantities that scare even the primary dealers.  The benefit of having large volume trade in the screens is the mechanism of price discovery, and volatility.  Scaring the dealers (the liquidity providers) enables things like short covering rallies...long liquidations, supply distributions, and the like.  The more volatility...the more price discovery...the easier for day traders to read what the larger market participants are doing...and thus...the easier to follow along.

To a certain extent, this is just a gripe....but today, for example, 10yr futures were stuck in a 4 tick range for 7 hours (10am --> 5pm)!!!  This is just not normal.   Also, the price discovery process (price action) normally has a volume measured moves based on a certain quantity of volume.  With the lack of significant volume trading in the market...price doesn't move enough...doesn't create enough volatility to get an accurate "read" on the direction of interest rates.  Volume tends to create volatility...which tends to bring in additional market participants.  Its that age old saying...volume begets volume.

This week, daily trading volumes were 50-60% of average for the past 2 months, and intraday volatility was reduced by a similar margin.  I'm hoping that the hot July in NYC is not a coincidence.  I'm hoping that many large traders are simply on vacation and will return at the end of the summer.  If not...then my trading strategy will need to change.

Of course, my complaining won't change anything.  Until trading volumes return, the best volatility will occur during London trading hours into the NY morning...and the NY afternoon will be the dead zone.

Monday, July 15, 2013

US Long End Auctions (10yr and 30yr) Post Mortem

Wednesday and Thursday of last week were the US 10yr and 30yr auctions.  These auctions (combined with the price action in the secondary market leading up to the auctions) are the best times to gauge demand for UST paper.

1st the 10yr:
Heading into the 1pm auction, the market for UST paper was weak.  The most recent comments out of the Fed indicated the FOMC was planning on tapering their bond purchases starting as soon as September.  The mkt grinded sideways to lower all morning.  The mkt went into the auction at the lows of the day (asof 1pm)....and the auction came at the market (0.1bp small of a tail lets call this on-the-screws).  The FOMC minutes (came out 1hour after the 10yr auction) indicated that some Fed Members wanted to taper QE back in June.  This surprised the mkt and the 10yr mkt traded lower to below the auction price by a few ticks on decent volume.

Then Bernanke spoke to an economic club, and around 4:45pm EST made comments that the recent labor market statistics, including the unemployment rate, under-represented the problems in the employment situation in the US.  Bernanke said that because of this, the Fed would remain accommodative for an extended period of time.  He didn't specifically clarify if this comment on "accommodation" was referring to the Fed Funds rate, or to QE.  The market knee-jerk reaction interpreted this comment as "Taper-OFF"(recall this was just 3 hours after the hawkish FOMC minutes)....and 10yr futures rallied  1 1/4 handles in a very illiquid session (from 125-05 to 126-15+).  This felt like another "game changer"  Many market strategists were unhappy that BB would make such comments when the mkt was for all intents..closed.

2nd the 30yr:
The price action for the 30yr bond auction the following day was interesting.  The 30y mkt almost retraced its entire BB spike from the prior day, for the setup going into the auction..the mkt was illiquid and low volume all day.  It seemed like the auction setup was the only trade occurring that day.  Again, going into the auction the mkt was trading at the lows of the day asof 1pm (recall, this is 18hrs after BB's comments the prior day).  This time, the auction came stronger and stopped short (came thru the 1pm price) by 1bp (about 5+ ticks on the 30yr).  The 30yr bond proceeded to rally 16 ticks from the auction into the close.  Many traders described this auction sarcastically as "couldn't see that coming."

In the 2 days since the 30yr auction, many strategists and Fed watchers have commented that the "Taper-OFF" reaction to BB's comments may have been "too artificial".  They point out that BB's comments regarding accommodation could just as easily be referring to the Fed Funds rate, and have no bearing on QE. (since traditionally, the Fed Funds rate was THE mechanism for accommodation).  BB must know this (he's not that clueless) many rates strategists describe his comments as an effort to talk the mkt "up" without needing to do very much.  The longevity of such verbal efforts seems to be decreasing.

The Fed and Fed reporters have stressed that they think they can "Taper" QE and have no impact on market interest rates.  The market has made an effort to say "incorrect."  While the Fed has explicitly stated that a reduction in QE could be followed by an increase in QE if they detect negative influences in the economy...the market just doesn't believe.  While its true the mkt is forward looking....they just aren't THAT forward looking.  The market sees any reduction in QE as a direct path to NO QE.

Nobody knows for sure if QE is actually helping main-street (arguments go both is usually the case in economics)...but we can say with confidence that QE is helping other asset classes such as the stock market and related interest rate markets (hello mortgages).

As i write this, the UST market is selling off and looks like it may "fill the gap" to BB's comments on Wednesday evening.  Regardless...the ultimate conclusion of the mkt thus far is one of confusion.  Will the fed Taper...or won't they.  Both the FOMC minutes, and the Fed Speak afterwards from various fed speakers have been contradictory.  Its clear there is a disagreement inside the Fed itself regarding what the best course of action is.

Over the past week, volumes in the US treasury market have been hovering below 70% of average.  There have not been large initiating trades one way or there other.  Perhaps this is due to the summer season surrounding July 4th (its gotten hot in NY...and many participants have gone on vacation).

I myself eagerly await the return of the large i can return to my usual tactic of following them (my most profitable trading strategy).  Until then, its probably gong to be a dicey market to trade.

good luck

-govt trader

Tuesday, July 9, 2013

Converting UST Market Volume to Interest Rate Risk

A question from a subscriber about Market Volume, and my response:
Govt - can you explain this: "12mm/bp traded over last 30 minutes". I know you are referring to volume, but beyond that, I can't figure out what it means! Also, it would be helpful (at least in the early stages until we catch on) to get some explanation for the meaning you are taking from the volume figures that you cite. They must be important to you, or you wouldn't be tweeting them. Is it high volume or low volume - and how is that volume figure supporting or invalidating your working hypothesis?
--my response-- (this is a little long..but you'll learn something)

 US Treasuries all share certain characteristics 
-they all pay their coupons semi annually
-they all accrue interest on the same Actual/Actual DayCount Schedule
-they all have their own maturity date
-the formula to calculate price --> yield...and yield --> price is the same for all UST securities (the only variable is coupon and maturity date)

So, we can do some simple bond math on them...and apply it the same way.

Now, all US treasuries have a market price...and each market price is equivalent to a specific yield. For example, today (July 9) the current 10yr notes bid price of 92-07 = a yield of 2.653%

There is a formula to calculate a yield from a bond, given its price.  There is a similar formula to calculate a bond price, given its yield (these are inter-changeable).

We can change the yield on any UST security by 1 basis point, and we can then calculate the new price (so back to our 10yr example 92-07 = If we change the yield to 2.643% (1bp lower)...we can then calculate the equivalent price...which would be approximately 92-095 (2 5/8 ticks higher in price for 1bp lower in yield). This is the sensitivity of the 10yr note price (2 5/8 ticks) for a 1 basis point change in yield of the 10yr note (from 2.653% to 2.643%).

If we do the same exercise for the other securities on the curve, we will find different price sensitivities to a 1 basis point change in yield:

2yr    5/8 of a tick
5yr     1+ ticks
10yr   2 5/8 ticks
30yr   5 1/4 ticks

--Update --> including treasury futures

ZF (FV)    1 5/8 ticks
ZN (TY)   2 1/2 ticks
ZB (US)   4 3/4 ticks
UB (UL)  7 1/2 ticks

These are the changes in price for these securities for a 1 basis point change in each of their respective yields. This value is referred to as PV01...or...the "price value of 1 basis point" change in yield for the security.

Why do we care about this?
Because the 1st assumption in bond trading is that all securities move in lock step on a YIELD basis. Meaning..when the market rallies in yield...and all securities rally or selloff in the same amount of yield. (of course we know this is not true...and we deal with that later...but it is a 1st level assumption which allows us to speak about bonds with different maturities in a apples to apples comparison)

So, imagine you are long 1mm 10yr notes...and you decide you do not want any outright exposure or market risk, to a change in the level of interest rates.

You have a couple choices:

1) you could sell your 1mm 10yr notes making you flat
2) you could sell another instrument (perhaps a 30yr bond), as a hedge against your 10yr notes

In option do you decide how many 30yr bonds to sell, to hedge your position in 1mm 10yr notes?

The answer comes from their PV01. You will notice that the 30yr PV01 (5 1/4 ticks) is DOUBLE the PV01 of the 10yr note (2 5/8 ticks). This means that the 30yr bond price will move DOUBLE the amount that the 10yr note price moves assuming parallel yield changes...(for ex...all UST yields increase by 4 basis points).

So...if the 10yr note sells off by 4 bps (remember, price down = yield up)...
the price change will be 4 * 2 5/8 = 10+ (10 and 1/2) ticks.

If the 30yr bond sells of by 4 basis points...its price will change by 4 * 5 1/4 =21 ticks.

In order for the P&L of the hedged position to be = 0 in this scenario (parallel yield change)...we would need to sell 0.5mm 30yr bonds to hedge our long 1mm 10yr notes position.

Or in other words..the hedge ratio = 10yr PV01 / 30yr PV01....or (2 5/8) / (5 1/4) = 0.5

We now know that 1mm 10yr notes are equivalent in a parallel yield move to 0.5mm 30yr bonds. The P&L of the 2 positions will be equal and offsetting.

I hope you are still with me...cuz we are just about done...

The actual dollar value of 1mm 10yr notes = $312.5/tick * 10yr PV01 (2 5/8 ticks) = $820.
This is referred to as DV01.  The only difference between DV01 and PV01 is multiplication

DV01  = PV01 (ticks) * $312.5 / tick / 1mm

This mean that when the 10yr note moves 1bp...a position of 1mm will move (in P&L) $820. A position of 2mm would move (in P&L) 2 * $820 = $1640

If we do the same exercise for the 30yr bond...we find that 1mm 30yr bonds PV01 of 5 1/4 ticks * $312.5/tick = $1640

So, 1mm 30yr bonds are equivalent to 2mm 10yr notes (in P&L exposure assuming a parallel move in yields).  

Back to quoting Market Volume 

When we count the volume traded in the mkt...we first count each security on its own (10mm 10yr notes trade....then 7mm 30yr bonds trade, etc...).

However...if we want to add the volume together (in this example 10 + 7 = 17)...we can't...because it wouldn't be an apples to apples comparison in terms of P&L exposure. While saying that 17mm US Treasuries just traded is technically does not give us a meaningful description of how much INTEREST RATE RISK just traded.

In order to get an apples to apples comparison..we need to convert each securities VOLUME into DV01 (dollar value of a 1bp move in yields). We do this by multiplying each securities VOLUME * PV01 = DV01

When i quote 17mm/bp traded in the last 30min..i am saying that the total DV01 traded = (the sum*product of the Volume * PV01 for all US Treasuries that have traded in the market in the specified time interval).

If i wanted, i could then convert that DV01 into an equivalent amount of 10yr notes.
17mm/bp = $17,000,000 (total DV01 traded) / $840 (DV01 of 1mm 10yr notes) = 20,238 million = 20.238 billion 10yr notes.

These are equivalent statements. 

By aggregating the DV01 traded of all the securities together...i'm not telling you what the composition of the market volume was. Perhaps all the volume came from trading 5yr notes...perhaps it was all 10yr futures (ZN)?? In that case, would it make sense to quote market volume in terms of 10yr notes? (while equivalent is technically can be misleading)

This is why i tend to quote DV01 traded (17mm/bp)...rather than 10yr note equivalents traded (20bln 10yr equivalents) (even though they are interchangeable)

In other words...when we are counting total market volume, we tend to not care what the structure or composition of the volume is (how much much much 2yr...etc..)...we just want 1 number..which makes it easy to grasp quickly what the "status"of the market is.

Now, of course its also valuable to know where the volume is coming from (how much DV01 from each instrument) when the curve steepens or flattens...or in other words...what is the composition of the volume trading in the market (how much much much 2yr...etc..) compared to the change in yields of the various instruments.

So in 1 sense...we don't care about composition..and in another sense...we do care.

When we just want to know how much volume has traded...we don't care so much...but when we want to understand the movement in the curve (when it steepens or flattens)...then we do care about the composition of the market volume....but that is a different type of question.

That is the end of today's lesson - are there any questions?

Monday, July 1, 2013

My response to a valid reader question - "why your premium service?"

RE: Marks comment

why sell the 'tips'? Why not just make a lot of money and retire? I am genuinely interested in your service, but as Livermore said "If you need someone to tell you when to buy, who will tell you when to sell?"

i completely understand this comment...and until i started blogging / tweeting my trade ideas over the last year...i had the exact same opinion of trading "advice" from an anonymous guy on the internet.  Here is my response. 

I don't intend to just give my buy/sell levels (yes, i'll give you those).  i don't think these alone will really help most traders tho...because of the inherent fear of trading which you allude to) premium service is more than 70% additional "education" info where i talk in more depth why i make certain decisions and how my thinking process works.  

My subscribers will attain their best trading results, as do i, when  they understand what type of pattern the market is experiencing.  Sometimes the pattern my model identifies is very short term (minutes to hours)...and sometimes slightly longer (hours to days).  Once  my model (and hence, i) understand the type of pattern that the market is going thru, my model creates buy and sell levels that make sense in light of the pattern.  If the market generates price action which invalidates the pattern, then the model gets stopped out and continues searching for the next pattern.

As time and days go on, subscribers to this service will learn to identify and classify these patterns.  Sometimes, an invalidation of a pattern can become a pattern in and of itself.

Over time, this will be similar to the experience that my junior traders used to have (for years, i was the guy on the desk that had the kid from MIT/Duke/Princeton sit next to me for 6-9 months as i taught / he learned the internal dynamics of the treasury market).  I have a technology background, so teaching was not hard for me, because i started out as an "outsider".

While i'm confident that if you were to follow all my trades and trade alongside me, you would have a positive net P&L, i'm even more confident that after spending 6 months to a year on my service, you will be able to anticipate many of my trades and ideas before i even post them...or when i post them you will think..."that's exactly what i was thinking"..and the reasons behind my thought process can help explain your own thought process, and give you more confidence, which will help you enter and exit the mkt (selling a pop...or buying a dip) instead of being paralyzed with fear and greed (the demon of most traders).  Its scary sometimes to sell a pop to initiate a short position (if the market is popping...then maybe it will keep going up...yikes!!!)  However, selling the right pops and buying the right dips are the ideal methods to generate P&L in the interest rates markets.  Generally, interest rates tend to mean revert so a fair value level.  The problem is that this "fair value level" changes from day to day...and the change can be a lot.

If you spend enough "screen time" staring at the treasury market along with me, i suspect you will start to notice patterns.  This normally takes at minimum, 1-2 years (don't take my word for it...go ask other traders).  It took me about 3 years before i felt like the treasury market was "speaking to me"...i was slow in this regard.   Part of recognizing patterns is having a frame of reference to start with.  Everybody has access to intraday price that is the common form of reference...but without a basic guide on how to read a price is difficult to discern the patterns at play.  The model that i've developed gives a frame of reference from which to view these price charts.

This "feel" that i refer to is not always there...but after enough time, i am now able to feel when the mkt is speaking to me...and when it is not (thanks to this model).  Now, when the mkt speaks to me..i trade..and when it doesn't i try to not trade.  You might say that the mkt has an ebb and flow like the ocean...and after enough time...when the waves get sufficiently really can feel them in the screens (the model discerns a pattern and projects a continuation of the pattern creating buy and sell signals).  One of the mistakes traders often make when they first experience this is to think that the "feeling" of a particular read will continue.  For example, after enough time...i've learned that most "ebb and flow feelings" are temporary...and are best used to get in and out of the mkt in short time-frames (usually less than an hour..and usually for 4 to 8 ZN ticks).  When i first experienced this..i was great at picking entry levels and my position would quickly make 3-5 ticks...but then the feeling would fade, and instead of exiting for a small profit when the "feeling" faded...i would wait for a large profit ("i think" would turn into "i hope") and instead i would watch the mkt take it all away.

This type of mkt feeling is difficult to ascertain if you haven't sat on the trading desk of a primary dealer, because of all the information that passes thru those desks (some info is trades..some is the opinion of very large traders such as Buffet, Soros, PIMCO, SAFE and the like talking with the head of the desk).  That flow information (while useful for scalping) is most valuable as a tool to learn what effect certain volume trades have on the mkt (you can't imagine the power and control you have over the mkt when you get to do these large flows and nobody else knows).  After seeing enough of can get a feel for how certain chart patterns combine with certain volume trading patterns (i've embedded these patterns into the model).

I would have never picked up this minutiae if i wasn't part of a primary dealer desk...its not fair..but that's just how it is.  Maybe there are trading prodigies out there who can learn these patterns without this experience..but i wouldn't count myself among them.

The "bonus" is that after having seen enough of these flows, how they impact the mkt and how the mkt reacts to the flows hitting the screens, i can now "see" the flows in the screens even tho i'm NOT at a primary dealer anymore...and i have a general feel for which customers do what...both types of trades...typical size..typical placement...etc.  I've taken this experience and built a trading model around it. "conclusion" service, while useful in itself for the trading more useful as an educational service...and that's how i'm billing it...a premium education service.  If you were trading equities...i'd suggest one of the equity Virtual Trading Floors...where experienced intraday equity traders talk about patterns they see in the mkt in a chat-room.  However, my experience is in US Treasuries.

If you want to trade US Treasuries, and you have the opportunity to get a spot on one of the primary dealer desks...i strongly suggest you take it.  There is no substitute.  My service is an option for those who don't have that opportunity.  I haven’t found another twitter service / chat-room or other opportunity to follow/ talk with a former primary-dealer US Treasury trader with my experience who has made themselves available to the internet..   If any of my readers are aware of one…I’d happily reach out and get in touch with them (because I enjoy the camaraderie). 

Of course i already speak with my former colleagues…but you can never have enough info when you are dealing with a market that has so few large participants compared to the size and depth of the mkt.

Who do you talk to when you are trying to understand the US treasury market?…could you trade better if you widened your community??

Regarding questions about this new service and my own trading...yes...every trade i post on the private feed is a trade that i am making myself.  I cannot tell you what to do....but i can tell you what i'm doing.  I can tell you where / when i buy / sell..and i can tell you why...and that is exactly what i do in the private twitter feed.