Bund/BOBL spread: Short Sale Parameter Revision

Trailing Along

For the last several months, the IMC blog has been stalking the Bund/BOBL spread for a short sale.  Basically, we’ve been trailing the spread with a contingency to get short on a break of a prior month’s low.

It hasn’t happened yet.

September Bund BOBL spread daily

September Bund BOBL spread daily

Trading for the month of July ended today.  The September Bund/BOBL spread finished with a new contract high.  A mid-month correction left an obvious price support area on the chart at the July low of 32.54.  Therefore, we are going to raise the short sale parameters to enter on a break of this correction low.

The Bund/BOBL spread is rocketing higher as negative interest rates in Europe continue to propel the treasury spreads.  Once the “rocket” runs out of fuel, though, a significant reversal is likely.

One way we may know that the fuel is gone is when the spread breaks below technical support on at least two timeframes.  The break of the mid-July low would do the trick on the daily timeframe.

On the weekly chart, the rising 30-bar Moving Average may be the trip switch to keep an eye on.  When the Bund/BOBL spread closed below the weekly 30-bar MA in 2013 and again in 2015, it continued to trend lower for months afterwards.

Bund BOBL spread weekly (30-bar MA)

Bund BOBL spread weekly (30-bar MA)

Therefore, a close below the weekly 30-bar MA for the first time since August of 2016 would confirm that a downtrend is in motion.  Once that happens, we will likely be positioned on the short side and looking to add.

Trade Strategy:

Cancel the current hypothetical order to short the Bund/BOBL spread and replace it with a new hypothetical order to sell one September Euro bund contract and simultaneously buy one September Euro BOBL contract if the spread closes below the July low of 32.54.  Initially, the spread will be liquidated on a two-consecutive day close 10 ticks above the contract high that precedes the entry (currently at 34.26). 

Loon/Kiwi Spread: A Long Position Was Initiated

Buying the Bottom…Hopefully

Today the IMC blog initiated a long position in the Canadian dollar/New Zealand dollar spread at 4.75 cents (premium Canada).  Initially, the position will be liquidated on a two-consecutive day close below 3.75 cents.

Canadian $ Kiwi $ spread daily

Canadian $ Kiwi $ spread daily

We are now monitoring on the 75-day Moving Average.  The last few closes above the 75-day MA have been followed by further gains of several hundred more basis points in this spread.  Therefore, a close back above the 75-day MA for the first time since the end of May would be a green light to add more.

However, instead of setting parameters to automatically add on the close above the 75-day MA, we will assess the situation after it occurs.  We will need to see how far above the 75-day MA the Canadian dollar/New Zealand dollar spread closes, whether or not it experiences a pullback, and what size of countertrend moves the spread makes in order to get a feel for the volatility.  That way, we will be able to make a better-informed decision on entry and initial exit levels.  This will also give us a better feel for the reward-to-risk potential on the trade.

Gold/Crude Spread: Verbatim Replay of 2015


The IMC blog reentered a hypothetical short position in the December gold/crude (x3) spread at -$14,430 (premium crude) on July 13th.  The position was then liquidated at -$2,620 (premium crude) on July 26th.

The mid-July short sale was made after the spread neared a Fibonacci .618 retracement, briefly closed above the 100-day Moving Average, and then rolled over.  The idea was that this should be the end of a bear market rally.

The position was liquidated because the spread broke out above the early July price peak.  This put back on track for a continuation of the bull market.

Right now, it appears that we are watching a rerun from last year.  Recall that the spread peaked in January 2015, closed below the 100-day MA in mid-April for the first time in several months, and closed back above the 100-day MA in early July.  This was followed by a continued bull run and new contract highs.

This year the spread peaked in January (again!), closed below the 100-day MA in mid-April for the first time in several months (again!), and closed back above the 100-day MA in early July (again!).

December Gold Crude Oil (x3) spread dailyIf the analog continues, we should be looking at new contract highs soon, an autumn pause, and then more upside right at the end of the year.  Until the 2016 pattern shows a divergence from the 2015 pattern, we are going to hang back on the sidelines.

Crude Oil/Nat Gas Spread: In the Trade…Now Let’s Add!

Picking a Bottom…Or Catching a Falling Knife

Today the blog entered a long position in the December crude oil/natural gas (x2) spread at -$17,000 (premium nat gas).  We’ll exit on a five-day close below -$20,000.

Usually, we don’t like to buy a spread when it’s plummeting.  The reason we took the chance on this one, however, is because it seems to find a bid down at these levels.  The spread finished at -$17,690 today.  As we noted before, the December crude oil/natural gas (x2) spread just hasn’t been able to stay below -$17k for more than a day or two over the last couple of years.  It’s been like trying to hold a helium balloon underwater.  So if you want to try picking bottoms, this is certainly the place to do it.

Pressing Our Luck

We also mentioned in the previous post that the last three drops below -$17k were followed by rebounds of nearly $16,000, approximately $13,500, and $13,300 over the following months.  So we’re looking for a sizable bounce to start soon.  Therefore, we would like to quickly add to the position if the spread does start to recover.

So far this month, the small bounces in the December crude oil/natural gas (x2) spread have ended either side of -$14,500.  Prior drops below -$14k have only lasted four weeks at the most.  Anybody who bought the spread at -$14k or lower and held on never had to wait for more than a month before the position was profitable.

December Crude Oil Natural Gas (x2) spread daily

December Crude Oil Natural Gas (x2) spread daily

Since the spread has already been below -$14k for just over three weeks, we can use a close back above this level as a signal that the bottom is in.  Furthermore, a close above -$14k would mean that the position we entered at -$17k is already showing an open profit.  This would allow us the cushion we need to add to the trade.

‘Add-On’ Trade Strategy:

For tracking purposes, the blog will make a hypothetical ‘add-on’ trade by buying one December 1,000 barrel crude oil contract and simultaneously selling two December 10,000 MMBtu (million British thermal units) natural gas contracts on a close above -$14,000 (premium nat gas).  Initially, this ‘add-on’ spread will be liquidated on a two-day close below -$17,000.

Soy Meal/Bean Oil Spread: The ‘Add-On’ Plan

Getting Aggressive

The IMC blog shorted the September soy meal/bean oil (x2) spread at +$500 (premium meal) on July 7th.  Prior bull markets that put the meal side of the spread at a premium were followed by bear markets that took the soy meal contract to a discount of at least $10k to a pair of bean oil contracts.

September Soy Meal Bean Oil (x2) spread daily (2)

September Soy Meal Bean Oil (x2) spread daily

As a matter of fact, a majority of the bear markets crushed the spread down to -$20,000 (premium bean oil) or lower.  So with the spread hitting a two-month low of roughly -$2,500 this morning, this means there’s still a lot of potential money to take off the table in this trade.  Therefore, the blog is going to pyramid this position by progressively adding more spreads as the price decline continues.

Setting the Intervals

As we explained in the post on the soy meal (x2)/soybean spread, a method I like to use for determining the ‘add-on’ levels or intervals is to measure the countertrend moves of the prevailing trend.  Then we set the intervals at a larger size than the countertrend bounces in order to give the spread the room it needs for countertrend moves without running a high risk of getting our exit parameters triggered.  In other words, we are measuring the volatility and trading around those measurements.

The September soy meal/bean oil (x2) spread peaked just three weeks ago.  So far, the largest countertrend bounce has been $1,116.  Given the fact that the spread has dropped nearly $2,500 over the last four days, the small size of the countertrend bounce seems suspect.  Therefore, I want to take a look at the countertrend pullbacks that the spread endured during the three-month bull run off the early April low.

Until mid-June, the largest pullback that the spread made was $1,362.  It finally saw a correction of $2,130 off the mid-June high and a final top was established a couple of weeks later.

So up until mid-June the countertrend moves maxed out below $1,400 and since the current peak, the countertrend bounce was just over $1,100.  This indicates that the volatility is not that great.  We’ll use that to our advantage and set the ‘add-on’ intervals at $2,000 to add to the position.  This is several hundred dollars more than the typical countertrend moves on the ride up and, so far, on the way down.

‘Add-On’ Trade Strategy:

The blog will sell another September soy meal/bean oil (x2) spread on a close below -$2,500 (premium bean oil).  If filled, exit on a two-day close of $2,000 or more above the entry price.

If the ‘add-on’ order is filled, sell a third September soy meal/bean oil (x2) spread on a close $2,000 below the entry price of the prior ‘add-on’ position.  If filled, exit on a two-day close of $2,000 or more above the entry price.

If the second ‘add-on’ order is filled, sell a fourth September soy meal/bean oil (x2) spread on a close $2,000 below the entry price of the most recent ‘add-on’ position.  If filled, exit on a two-day close of $2,000 or more above the entry price.

To enter a short position in this spread, sell one 100-ton September soy meal contract and simultaneously buy two 60,000 lb. September bean oil contracts.

Meal/Bean Spread: Build the Pyramid

Full Court Press

On June 23rd the IMC blog initiated a short position in the Dec-Nov soy meal (x2)/soybean spread at +$21,125 (premium meal).  The spread touched a two-month low this morning, so the trade is making some headway.

Prior tops at +$20k or higher (premium meal) were followed by drops back under +$9k (premium meal).  Therefore, we fully expect the Dec-Nov soy meal (x2)/soybean spread to return to the April 4th contract low of +$8,975.

December Soy Meal Soybean (x2) spread dailyFurthermore, prior to the 2012 drought, tops established north of +$20k were followed by bear market declines to ‘even money’ or lower.  So a break of the April low could indicate that the Dec-Nov soy meal (x2)/soybean spread has further downside potential of several thousand more dollars.

Given the fact that the spread seems to be in a fast decline and our minimum downside price expectation is several thousands of dollars away, it makes sense to pyramid this position to take full advantage of the decline.

Setting the Intervals

We’re going to add more short positions as the spread declines.  The ‘add-on’ positions will be entered in intervals as the down trend continues.  The trick, of course, is to know where to set those ‘add-on’ intervals.

One method I like to use is to measure the countertrend moves of the prevailing trend to get a sense of how much room we need to give the spread.  Then set the intervals at a size that is wider than the countertrend moves.  That way, we are allowing the market to have the breathing room that it has already demonstrated that it needs.

Since peaking on June 10th, the largest countertrend bounce in the Dec-Nov soy meal (x2)/soybean spread was the $1,970 rally off the June 24th low.  Therefore, the blog will use intervals of $3,000 to add to the position.  This is more than 50% bigger than the largest countertrend bounce to date, so the spread should have plenty of elbow room.

‘Add-On’ Trade Strategy:

The blog will sell another Dec-Nov soy meal (x2)/soybean spread on a close below +$18,125 (premium meal).  If filled, exit on a two-day close of $3,000 or more above the entry price.

If the ‘add-on’ order is filled, sell a third Dec-Nov soy meal (x2)/soybean spread on a close $3,000 below the entry price of the prior ‘add-on’ position.  If filled, exit on a two-day close of $3,000 or more above the entry price.

If the second ‘add-on’ order is filled, sell a fourth Dec-Nov soy meal (x2)/soybean spread on a close $3,000 below the entry price of the most recent ‘add-on’ position.  If filled, exit on a two-day close of $3,000 or more above the entry price.

To enter a short position in this spread, sell two 100-ton December soy meal contracts and simultaneously buy one 5,000 bushel November soybean contract.


ULSD/Crude Spread: Good Place For a Trend Change

A Matter of Semantics

I’ve been trading futures for twenty-five years now.  As on old guy, I still catch myself referring to one of the crude oil derivative contracts as heating oil.  But it’s not.

To comply with EPA rules, the specs for sulfur content on the heating oil contract was changed from 2,000 PPM (sulfur content) to 15 PPM.  Due to this significant modification, the contract was renamed the ultra-low sulfur diesel fuel (ULSD) contract.

Now when you ‘crack’ crude oil at the refinery, the two resulting products are gasoline and ultra-low sulfur diesel fuel (ULSD).  But forgive me if I still occasionally call it the heatin’ oil.  I’m an old trader.

Still Connected

Despite the change in the contract specs, ULSD and crude oil are still highly correlated.  So it’s still a viable candidate for spread trading.

ULSD Crude Oil overlay (nearest-futures 2013-2016) weekly

ULSD Crude Oil overlay (nearest-futures 2013-2016) weekly

As a matter of fact, look at the chart of weekly closing prices of ULSD and crude oil since the contract specs were changed in May of 2013.  Can you tell which market is which?  Didn’t think so.  The ULSD price plot is the red line and the crude oil price plot is the black line.  These two markets appear to stick together like Siamese twins.  As a spread trader, this is exactly what I like to see.

Historic Extremes

Using the heating oil data up until 2013 and the ULSD data since then, we can get over three decades of price history on the charts.  The heat/crude spread used to be expensive when it reached six dollars or higher.  That ceiling was permanently broken in 2005.

In 2004, we learned that crude oil’s previously unsustainable 1990 Gulf War high around $40-per-barrel could not only be surpassed, but in a few years consumers would wish the market could someday come back down to forty dollars!

By 2005, the heat/crude spread had blasted its way to new record highs as well.  The spread never made it back down to the six dollar area until the 2008 financial crisis caused a collapse in energy prices.  Even then, the spread oscillated around six bucks for several months in 2009 and took off on a run for record highs again two years later.

ULSD Crude Oil spread (nearest-futures) weekly

ULSD Crude Oil spread (nearest-futures) weekly

The heat/crude spread reached a crescendo at the October 2012 record high of $43.83.  After converting to the ULSD/crude spread in May of 2013, the spread has made runs to $35.13 in 2014 and $37.89 in 2015.  Despite the recent historic bear market in energies, the ULSD/crude spread has still never touched six dollars yet.

After walking through this history, this spread doesn’t seem very mean-reverting.  So what’s a trader to do?

I think there are two things we can do.  First, we should look at the ratio between ULSD and crude to normalize things.  This will compensate for the spread widening during the period of record high prices and tell us if the relationship between the two markets still shows mean-reverting tendencies.

Second, trade the price action.  Even if the ULSD/crude spread is not as mean-reverting as we might hope, the price action is still what traders should follow.  As long as we are following the trend, the “path or righteousness”, the path of least resistance, etc., we are on the right side of the market.  And if the trend lines up with the direction of historical mean reversion, then that’s just all the better!

Drum Roll…

Good news, folks!  The ratio between ULSD and crude oil does indeed show that the relationship is still very mean-reverting.  As was the case with the heat/crude ratio, the ULSD/crude ratio indicates that things are both expensive and temporary when the ratio exceeds 1.45:1.

ULSD Crude Oil ratio (nearest-futures) weekly

ULSD Crude Oil ratio (nearest-futures) weekly

When the heat/crude ratio reached 1.45:1 or higher in the past, it was only a matter of time until it dropped back down to 1.1:1 or lower.  Now that we’re dealing with the ULSD/crude ratio over the last three years instead, it appears that the upside deviations are followed by declines to either side of 1.2:1 instead of 1.1:1 like the heat/crude did.  This slightly higher downside destination likely just coincides with the change in contact specs.

Daily Pattern

The IMC blog is tracking the December ULSD/crude spread for a trading opportunity.  Now, the ratio is currently around 1.31:1.  Therefore, it is not at an historic extreme of any sort.  The ratio for the December 2016 contracts hasn’t been anywhere close to expensive since it reached 1.43:1 eleven months ago.

However, the spread is worth taking a look at.  First of all, the December ULSD/crude spread rallied as much as $5.90 from the January low.  It just recently closed just above the ever-popular 200-day Moving Average for a few days. Then it lost the upside momentum and started backing off.

This is somewhat similar to what happened last year.  First, the spread rallied as much as $5.81 from the January low.  In May it closed just above the 200-day Moving Average for a few days. Then it lost the upside momentum and started backing off.

Last year, the failure to make it clearly past the 200-day MA was followed by a multi-month decline of over fifteen dollars ($15,000) per spread.  This time, we’re starting from a lower price point.  But you can’t rule out a return to the January low or even another bear market leg down.

December ULSD Crude Oil spread daily (200-day MA)

December ULSD Crude Oil spread daily (200-day MA)

In addition, the multi-month high that the spread reached in early July happens to be just a dime past the Fibonacci .382 retracement of the entire decline from the 2015 high to the current 2016 low.

Over the last couple of months, the December ULSD/crude spread has been supported by the 100-day Moving Average.  The spread first cleared this technical barrier in mid-April.  A sharp pullback in early May caused a closed below the 100-day MA, but it was a one-day event.  The spread snapped right back and made its way to new multi-month highs where it finally reached the 200-day MA.

Daily Pattern

That brings us to what’s going on now.  From the July 11th high -where the spread tangled with the 200-day MA and the Fibonacci .382 resistance line- the spread sold off for five consecutive days and closed at the lowest price in a month.

The rally on Monday and Tuesday then took it back up to just three ticks shy of the Fibonacci .618 retracement of last week’s sell-off.

December ULSD Crude Oil spread daily (100-day MA)

December ULSD Crude Oil spread daily (100-day MA)

The July 18th pullback low at $14.01 and the rising 100-day MA (currently around $13.82) set near-term price support.  A break below this level could indicate that the early July test of resistance was the end of the multi-month run and that the recent bounce to a Fib retracement was a failed recovery.  If so, it would be a good technical reason to short the December ULSD/crude spread.

Trade Strategy:

The blog will work a hypothetical order to sell one 42,000 gallon December ULSD contract and simultaneously buy one 1,000 barrel December crude oil contract on a close below the 100-day MA (currently around $13.82).  If filled, risk a two-consecutive day close 50 cents above the highest close after July 18th (currently at $14.86).


10-Year Treasury Spread: Canada vs.US!

Close Correlation with Canucks

The correlation between the Canadian 10-year bond and US T-notes is strong.  I mean, really strong.  As a matter of fact, whenever someone gets busted trading US treasuries on inside information about a jobs report or something like that, you’d wonder why they didn’t just trade the Canadian bond instead of the T-note in order to stay off the radar.

By the way, I am not actually advocating this.  If someone does this and gets caught, don’t blame me!

Canadian bond T-note overlay weekly

Canadian bond T-note overlay weekly

There are periods where the trends in these bonds can diverge for a bit, but it doesn’t seem to last more than a few months.  So any diversion could be a setup for a mean-reversion trade.

There are also periods where one treasury will move faster than the other.  If this pushes the spread to historically extreme levels, you have another potential opportunity for a reversal trade.

Way Up North

Don’t forget that treasuries prices trade inversely to the interest rates.  So the global bull market in treasuries means we’re in a global bear market for interest rates.

The Canadian economy is weaker than the US.  Therefore, their rates are lower, which means that their corresponding treasuries are higher.

In the cash market, the yield on the US 10-year note is at 1.57% while the yield on the Canadian 10-year bond is at 1.10%.  That may not seem like a big deal, but the 42% percent premium on the US yield has pushed the Canadian 10-year bond futures contract to a full 16-point premium over the US T-note futures contract.  This is a rich premium.

At the peak in 2006, the premium on the Canadian bonds was less than half of what it is today.  When the 2006 high was surpassed two years ago, the spread between the two treasuries more than doubled over the next several months.

Canadian bond T-note spread weekly

Canadian bond T-note spread weekly

The only other time that the premium on Canadian bonds has been this fat or higher was for a few months either side of the Y2K (non-)event.  By the end of the year 2000, Canadian 10-year bonds were trading at a discount to the US T-notes.

Suffice it to say, the current premium on the Canadian bond/T-note spread is at an historically high level.  This could create a breeding ground for a major reversal.

The Wall

In February 2015, the nearest-futures Canadian bond/T-note spread cleared the 16.50 mark.  It peaked out at 16.90 less than a week later and reversed on a dime.  The spread then plunged over six points over the next two and a half months.

In August 2015, the nearest-futures Canadian bond/T-note spread once again clipped the 16.50 mark.  It peaked out just one day later and started a two-month decline.  The drop ripped nearly four and a half points off the spread.

Canadian bond T-note spread (nearest-futures) daily

Canadian bond T-note spread (nearest-futures) daily

Just a week ago, the nearest-futures Canadian bond/T-note spread closed above the 16.00 level.  This is not quite the 16.50 mark that told us to set the timer for the reversal the last two times, but it’s getting close.  Based on the daily chart for the September spread, there’s a good reason to start monitoring the situation closely.

Will the Fib Stop Cause a Flip-Flop?

The September Canadian bond/T-note spread faces technical resistance at the Fibonacci .618 retracement of the entire decline from the January 15th contract high to the April 29th multi-month low.  That Fib level is located at 15.97.

September Canadian bond T-note spread daily

September Canadian bond T-note spread daily

Last Wednesday, the September Canadian bond/T-note spread closed slightly above the Fib retracement for one day.  It then backed off and has been sitting still for the last week.  Traditionally, the Fibonacci .618 retracement is an ideal point to look for a secondary lower top to form before a market starts another leg down.  If this happens in the spread, we’ll take a crack at the short side.  If not, we will simply trail it higher with moving short sale parameters.

It’s also a point of interest that the spread closed about a third of a point below the Fibonacci .618 retracement last month and then endured a correction.  The June 2nd high set price resistance that was surpassed on July 11th.  The fact that the spread is now back below the early June top is indicative of a breakout failure, or what we like to call a Wash & Rinse sell signal.  If prices continue to erode from here, a short sale could be warranted.

Trade Strategy:

Place a hypothetical contingency order to sell one September Canadian 10-year bond contract and simultaneously buy one September T-note contract if the spread closes below 14.93 (premium Canada).  If filled, liquidate the position on a two-consecutive day close .20 points above the rally high that precedes the entry (currently at 16.05).

Scale Trading: Alternative Way To Trade Spreads

Dollar Cost Averaging Spreads a.k.a Value Investing In Commodities a.k.a Trading Like Evel Knievel

Here’s something coming out of left field for spread traders.  It’s an idea I’ve been pondering for some time, so I’ll go ahead and spill it in this post.

I started my career at a commodities firm in 1991.  At the time, it seemed like commodities could only go in three directions: sideways…down…and down even more.  I can definitely tell you that learning how to trade commodities -an asset class traditionally favored during an inflationary environment- during a prolonged deflationary environment will certainly teaches one how to play good defense.

One trading strategy that became popular with our largest clients at the brokerage firm was scale trading.  Some guys paid several thousand dollars to attend a weekend seminar in Europe that taught them how to do it.  Other guys spent a whopping $20 or $30 to buy a book called You Can’t Lose Trading Commodities that taught them the exact same thing.

The gist of the method was to find a commodity that was trading in the lowest 10% of its historic trading range (20 years or more) purchase contracts in progressively lower intervals (“scale in”), and sell the contracts at progressively higher intervals (“scale out”) for a modest profit on each contract.

Commodities Only

There are a few important rules that you absolutely had to adhere to.  First, scale trading was only to be done with futures contracts on tangible, consumable commodities.  The reasoning was that a commodity could never be priced at zero –like a stock or a currency- because it would always have some sort of intrinsic value.  A commodity cannot go bankrupt like a stock nor have a radical and permanent devaluation like a currency.

The lower the price, the more likely it was that the commodity was to an absolute bottom.  This is because supply would (theoretically) get restricted as suppliers cut back, demand would start to increase as buyers would stockpile inventory, and producers would even tap out if prices got near or below the cost of production.  Ultimately, the market had to turn around at some point.

The Deadliest Catch

As many traders in the ‘90s found out the hard way, there is a way that your account can suffer as though commodity markets actually can go to zero…or even lower.  That catch is the carry charge.

Carry charge is the storage costs, interest charges on borrowed funds, insurance, and other related costs on a commodity.  This is why the longer-dated futures contracts are normally trading at a higher price than the closer-dated contracts.

For example, on Monday the August crude oil contract closed at $45.24 and the November contract closed at $47.22.  So the November contract is price at a premium of $1.98 (4.3%) over the August contract because it is being stored for three months longer.

The February 2017 contract closed at an even higher price of $48.92, which is $3.68 (8.1%) more than the August contract.  This is six months further out.

Notice that the carry charge annualizes at a little over $7-per-barrel, which is a whopping 16%.  So if you own the August crude oil contract, keep rolling it over for the next year, and the nearest-futures price of oil does not increase, then you will lose seven dollars of 16% on a market that does not even go anywhere!

This is the hidden danger to scale traders.  To combat this, you have to scale trade the contracts that are several months out until delivery and have the price intervals set tight enough to generate enough sales to compensate for the carry charge.  We’ll discuss the intervals in a bit.

Better Be Capitalized

Get this: You don’t use any stop loss orders in the strategy!  Instead, you hold the losing positions and add even more when the market goes against you.  It’s similar to dollar cost averaging.  Therefore, the second rule was that you had to anticipate this and have a huge bankroll to scale trade in a market.

Let’s say you were buying crude oil contracts every dollar lower and you started at $20-per-barrel.  Suppose the price dropped an additional 25%.  By the time the market hit $15 and you bought another contract, you were behind a total of $15,000 on all of the prior contracts that you had purchased at $20, $19, $18, $17, and $16.

Another one dollar drop to $14-per-barrel and your drawdown would increase another $6,000 to $21,000.  This is because you would now have six contracts that you are behind on…and you are still buying more!

Take Some Off the Table

The next rule was that you had to sell each contract when it hit a predetermined profit objective.  Once that happened, you would place an order to buy it back at the original entry price.

This is where scale trading is different from dollar cost averaging.  Both strategies have you buying in intervals as a market moves against you.  With dollar cost averaging, however, you don’t intend to liquidate any of your holdings once the market turns around.  It’s usually a long-term strategy that’s often part of a Buy & Hold strategy.  With scale trading, the trader is trying to take advantage of the markets fluctuations by buying and selling.

Suppose you had 50-cent profit objectives on the crude oil scale.  The contract you bought at $20 would be sold if it hit $20.50, the contract you bought at $19 would be sold if it hit $19.50, etc.  So each contract would be up for sale $500 better than the purchase price (each contract controls 1,000 barrels of oil, so a 50-cent move = $500).  As soon as you sold the $19 contract at $19.50 you would immediately place an order to buy again if the market dropped back to $19.

Determining the Price Intervals

When deciding how close together the purchase and sale levels for your scale should be, you have to consider two things: Bankroll and market volatility.

The tighter together the purchase points are, the more contracts you will possibly accumulate and the more money you will need to back it up.  A scale that is setup to buy crude oil every 50 cents lower is going to need double the capitalization of a scale that is setup to buy crude oil every dollar down.

The volatility tells you how much the market swings around.  If crude oil often makes 50-cent fluctuations a wise scale trader will set profit objectives of 50 cents or less.  This will generate more activity.  If the trader uses profit objectives of two-dollar intervals instead, the scale would not be nearly as active.

Shooting for a two dollar profit instead of a 50-cent profit when the market is only giving out 50-cent profits can mean little or no trading activity.  This means there are no profits being generated that will pay down the carry charge.  If this goes on too long the impact of the carry charge will felt after a few rollovers and the trader could be down double digits in an inactive position.

My personal belief is that the volatility trumps bankroll.  A scale trader should determine what size price intervals are appropriate based on the market volatility and then determine the correct bankroll to trade it.  If the bankroll is beyond the trader’s capacity, skip the scale trade campaign for that market and look for the next one that does meet both the volatility and bankroll criteria.

Don’t Get Caught Short

Here’s a “set in stone” rule for scale trading commodities: Never, never, never scale trade a market from the short side!

The foundation of scale trading is built on the idea that commodity prices cannot go to zero.  On the flip side, there is no guaranteed limit to how high prices can go.  So it is impossible to be fully capitalized for a short sale in a bull market run that has theoretically unlimited upside.  It may work a time or two, but it only takes one runaway market to rob you of your wealth.

Let me give you an example of how that could happen.  Suppose you started shorting crude oil every dollar higher once it reached $30-per-barrel in 2000.  Back then you might have assumed that $40 was the max it should go.  After all, it barely poked its head above $40 during the Gulf War in 1990 and collapsed into the teens just three months later.  But just to be safe, you might have capitalized for a run to $60, which is 50% above the Gulf War high.  No way oil could ever get that high, right?  The global economy would collapse before then.

As it turns out, oil reached a peak of $37 in September of that year.  It was back down under $26 by the end of the year and trading below $18 a year after that.  You would have made a nice profit.

A year later in September 2002, oil hits $30 again and you do the same scale.  The market runs to nearly $38 over the next six months.  It finally breaks and drops to nearly $25 in April of 2003.  You’ve made another pile of money.  Since you are capitalized for a run to $60, you did not get too deep into your reserves and you never broke a sweat.

Two months later, crude pops above $30 again.  It traded both above and below $30 every single month for the rest of the year.  It feels like your printing money as you keep pulling profits out of the market during this choppy period.  What color of Mercedes will you buy your wife for Christmas this year?

When 2004 begins, oil is elevated to nearly $36-per-barrel.  You’re not the least bit worried.

Four months later, oil breaches the 1990 Gulf War high.  This is record territory and you have a ton of contracts on.

In the autumn of 2004, oil rips through the psychological fifty-dollar barrier and rockets to just above $55-per-barrel by October.  Now you should be sweating!  You’re capitalized for a run to $60.  It’s getting awfully close.

Relief comes in the form of a Q4 break to just below $41.  Since you kept scaling up, several contracts were cashed out with profits on the price break.  You congratulate yourself for sticking to your guns.

Then 2005 rolls around.  The market surpasses the October 2004 record high in March.

Four months later, crude oil hits and even breaks the $60-dollar mark.  Game over.  All of your money is gone.  One bad run and everything you made in the prior scales in this market is evaporated.

Even if you stockpiled all the winnings…even if you recapitalized for a run to $100…it doesn’t matter.  Crude oil hit $147 by the summer of 2008.  Nobody shorting this market on a scale up basis could have survived that.

One Exception…Kinda

Like any rule, though, there actually is an exception.  You could scale trade the interest rate products from the short side.  We actually did that for a while in the 90s.

The reason for this exception is because the price of interest rate products are inverse to interest rates, so shorting the futures contracts is really like “buying” or scale trading the long side of the actual interest rates.  And anyone who ever took a class in economics knows that interest rates could never go to zero.  That’s absurd.  Banks would be giving out free money at that point.  Ain’t gonna happen.

As we all know now, interest rates can go to zero.  Heck, we currently live in a world where interest rates go negative.  Who woulda thunk it twenty-some years ago?!  Best stick to tangibles like commodities for scale trading.

I will admit, though, that I am tempted to scale trade the short side of the Swiss 30-year bond.  The yield is negative right now, but I just have a feeling that Swiss interest rates could poke their head above zero sometime between now and the year 2046…

Drive Your Accountant Cray-Cray

When you trade commodities, the accounting is done on a FIFO a basis.  This “First In, First Out” procedure means that, if you own multiple contracts of a commodity, when you sell a portion of them they will offset the sale with the oldest position.

When we scale traded back in the ‘90s, we used LIFO (“Last In, First Out”) accounting to track our trades.  This is because we sold our lowest-priced contract first, which means the most recent one that was purchased.  So we had to call the accounting department at the clearing firm for every single trade that we wanted offset this way.  It helped us clearly track the scale trades, but the accounting department sure didn’t like us.

Scale Trading Spreads

So I got to thinking, “What if a trader scaled into a commodity spread that was near or at historic lows?”  The reasoning is that the spread is mean-reverting and it’s already at an historic extreme.  Therefore, it’s a strong candidate for an inevitable recovery.

In addition, I would want to target a spread that does not have seasonal patterns price in like grains, livestock, etc.

Furthermore, I would pick a spread that has a very low carry charge.  That way, being stuck in the scale for a prolonged period of time would not crucify me if I have to endure several rollovers.

In light of this, the blog is going to run a little experiment.  In addition to the normal spec trades, we are also going to run a scale trade in the copper/gold spread.  Better buckle up for this one!  Oh, and bring some money, lots of money.

The Gold/Copper Spread Scale

We’ve established in previous posts that copper and gold are highly correlated and that the spread between the two metals is extremely undervalued.  Nonetheless, here are a few highlights worth noting before venturing out into the deep waters with a scale trade campaign.

Copper (x2) Gold spread weekly (-$10k interval)

Copper (x2) Gold spread weekly (-$10k interval)

First of all, the copper(x2)/gold spread has been a screamin’ buying opportunity whenever it has dropped to -$10k or lower.  At this level, two copper contracts are valued at a ten thousand dollar discount to one gold contract.  The spread is currently sitting below -$21k.  In addition, just a couple of weeks ago the spread breached the weekly 2009 financial crisis low by a few hundred dollars and closed at -$29,915.  This puts the spread on record as the lowest price in history.

Second, the ratio between the value of the sum of two 25,000 lb. copper contracts and the value of one 100 oz. gold contract is historically cheap when it drops below 1:1.  At that point, it means that two copper contracts are worth less than one gold contract.  Recall that we look at the ratio to normalize a spread relationship in order to help us determine if the spread is truly priced at an historic extreme.

Copper (x2) Gold ratio monthly

Copper (x2) Gold ratio monthly

Furthermore, the ratio closed at a multi-year low of 0.78:1 on a weekly basis at the start of the month.  This level nearly matched the weekly 2009 financial crisis low of 0.71:1, which was the lowest close since 1987.

Now hold on to your hats for this one…there’s almost no carry charge to speak of.  As a matter of fact, you could even say that the carry on this spread is inverted!

Wait.  What?!

That’s right, ladies and germs.  The July-Aug 2016 copper(x2)/gold spread closed at -$20,155 and the December 2016 copper(x2)/gold spread closed at -$20,440.  That’s $285 lower.

And if you go out a year further to the December 2017 copper(x2)/gold spread, it closed at -$21,135.

What’s that saying about a gift horse?

Trade Strategy:

So here’s how we’re gonna play it:

The scale will be done on the December 2017 copper(x2)/gold spread.  Notice this is Christmas of next year.  The contracts are not up for delivery for nearly a year and a half.  So we won’t be talking about rollovers anytime soon.

The scale will be run with $5k intervals.  This means we will buy a spread every $5k down and sell it for a profit out every $5k up.

In addition, the intervals will be done on a closing-basis.  No intra-day trading going on here.  This not only means that we don’t have to watch stuff on a monitor all day, but it also means that we have the opportunity to get bonus fills on the entries and exits.  Any close that’s better than the actual interval (i.e. if we want to buy a close at -$15k and the spread closes $1,500 lower at -$16,500 or if we want to sell a close at -$20k and the spread closes $750 higher at -$19,250) is considered a bonus fill.

Furthermore, we will start the scale at -$5,000 (premium gold).  That way, the top spread position is liquidated whenever the spread closes as ‘even money’ or better.  This is where the sum value of two copper contracts is equal to or more than the value of one gold contract.  This puts the ratio between the value of the sum of two 25,000 lb. copper contracts and the value of one 100 oz. gold contract back up at 1:1 or higher.  When the ratio is above this level, we do not want to be in any scale positions.

Now, since the December 2017 copper(x2)/gold spread closed at -$21,135 today, we should already be holding four spread positions in the scale.  This is because it closed below the -$5k, -$10k, -$15k, and -$20k intervals.  Therefore, the blog will buy the four positions at today’s closing price and consider them bonus fills.  Based on the $5k profit objectives, they will be sold on closes above -$15k, -$10k, -$5k, and even money and reentered on closes below the corresponding interval. 

The next purchases occur on closes below -$25k, then a close below -$30k, and so on.  And just to make sure we are clear, one spread consists of two December 2017 copper contracts and one December 2017 gold contract.

Good luck and good trading!

Platinum/Gold Spread: Current Observations

Taking a Beating

A year and a half ago, the platinum/gold spread inverted. Historically, platinum has never stayed priced below gold.  It’s always been a good buying opportunity.  So the IMC blog jumped at the chance to buy.

As readers know, we’ve been in and out of this spread a few times over the last few years.  It feels like we’ve done nothing but get beat like a proverbial “rented mule” whenever we’ve bought the breakouts in the platinum/gold spread.  Despite the abuse, we are still fully expectant that the platinum will once again trade at a premium of gold.

Platinum Gold spread (nearest-futures) monthly

Platinum Gold spread (nearest-futures) monthly

It is important to remember that commodity spreads are mean-reverting.  History shows that the more extreme a trend gets –both in terms of price and time- the bigger the inevitable reversal.  This current bear market is certainly one of the most extreme that the platinum/gold spread has ever experienced.  Just take a look at the current stats:

-On June27th the nearest-futures platinum/gold spread closed at -$343.30 (premium gold).  This was a new all-time low.

-On June27th the nearest-futures platinum/gold ratio matched the January 20th low of 0.74:1.  This is the lowest that the ratio has been since 1982.

-The spread inversion has now lasted for one year and six months.  If the platinum/gold spread doesn’t rally over $231/oz. in the next four weeks, this duration will tie with the record inversion of one year and seven months that occurred from September 1981 and April 1983.

Moving the Moving Averages

In the spring of 2015, we used a breakout above the declining 50-day Moving Average as a trend change signal to enter long positions in the platinum/gold spread.  That didn’t work out so well.

Then we slowed things down and changed the parameters for our trend change signal to a breakout above the declining 75-day Moving Average.  The spread did not comply with this criterion, either.

After the 75-day MA sucker-punched us at the start of this year, we backed off even more and went with the declining 100-day Moving Average as the new standard.  Three months ago, the platinum/gold spread made a two-day close above the 100-day MA for the first time since the summer of 2014.  For the next few weeks, it looked as though our third moving average parameter was the right pick.  The third time was really the charm!

Well, that idea went to hell in a handbasket.  The spread crashed in the second half of June and knocked out the speculative positions with a loss.

Schools of Thought

Right now, when it comes to the platinum/gold spread, there are two opposing philosophies competing for our loyalty.  This first can be summed up with such clichés as “Staying the course”, “Sticking to our guns”, “Winners never quit”, “Always follow the system”, etc.  This school of thought would have us continue to take buy signals in the spread without pause and without question.

The second philosophy is summed up by phrases like “You got to know when to fold ‘em”, “It’s different this time”, “Doing the same thing over and over and expecting different results is the definition of insanity”, “Know when to pull the plug”, etc.  This school of thought tells us to abandon the idea of buying this spread since it has not yet worked out profitably over the last year and change.

The thing is, I don’t think this has to be a binary Yes or No decision.  My thinking is that we can stay the course by continuing to look for buying opportunities/signals.  After all, the history of mean reversion is on our side.  And it’s not like the ratio hasn’t been this low before and the inversion hasn’t lasted this long.  Even if the ratio and duration of the inversion posts new records, the mean reversion idea argues that a reversal is all the more likely.

At the same time, we can drop the signals that are not giving us the results we want and either adjusts the parameters or even use different trade signals altogether.

This is actually what the blog has been doing all along.  We have slowed the moving average parameters after each signal failure.  I offer the analogy that it’s akin to tuning in the dial to get the exact frequency of a radio station.  If the 50-day MA is giving us “static”, then we use the 75-day MA, if the 75-day MA still doesn’t provide a clear signal, we switch to the 100-day MA, and so forth.

Parameter Recalibration

Since the 100-day Moving Average failed us on the last go around, we are now going to cut the speed of the moving average in half.  This means we are now going to use the widely-followed 200-day Moving Average to tell us what the trend is.

Legendary trader Paul Tudor Jones said, “My metric for everything I look at is the 200-day moving average of closing prices.”

This multi-billionaire made his fortune from trading commodities.  So hopefully looking at one of PTJ’s favorite metrics to determine the trend means that we’re in good company here!

Platinum Gold spread (nearest-futures) daily with 200-day MA

Platinum Gold spread (nearest-futures) daily with 200-day MA

On the nearest-futures daily chart, the platinum/gold spread made a two-month rally into the first half of May.  The rally ended after the spread tapped the 200-day MA and backed off.

Last week the spread finally made a two-day close above the 200-day MA for the first time in exactly twenty-three months.  So maybe…just maybe…this is the real deal?!

We’ve been burned a time or three by chasing the breakouts above various moving averages in the platinum/gold spread.  Not by a breakout above the 200-day MA, though.

However, the spread has already rallied $112/oz. over the last three weeks.  The last time it rallied anything close to this was the $107/oz. rally off the early March low.  And that one took two months to play out.  So it’s feasible that the spread has run a little too far, a little too fast.  It may be vulnerable to a pullback.

Ratio Development

It was mentioned earlier that the nearest-futures platinum/gold ratio matched the January 20th low of 0.74:1 in late June.  Now that it is closing in on 0.83:1, it appears that a double bottom is being established between the January 20th low of 0.74:1 and the June 27th low of 0.74:1.  All it needs now is a close above the May 2nd high of 0.84:1 to confirm it.

Platinum Gold ratio (nearest-futures) daily

Platinum Gold ratio (nearest-futures) daily

As an interesting observation, a double bottom in the ratio to mark the end of the two-year bear market would be the mirror image of the double top in the ratio that marked the start of the bear market.  Furthermore, the double top was established in January and June of 2014, while the current double bottom was established in January and June of 2016.  How cool is that?!

We Can Dream

An ideal scenario would be for the platinum/gold spread to clear price resistance at the early May 5th multi-month high of -$208.50 and then make a pullback near that level and hold.  It may offer an attractive enough reward-to-risk ratio to take a shot at it on the pullback.  If it works out, the position could even be quickly pyramided by adding on a breakout above the high that precedes the pullback.

Now that the important 200-day MA has finally been surpassed…and the three-week run looks extended at the same time…we will vigilantly monitor how the situation unfolds.