The 144-day moving average of gold, as discussed in the comments of an earlier post, appears to be significant: it's been the lower bound for the price of gold since early 2009. Incidentally, contrary to what I thought, the number 144 isn't quite random - it's a Fibonacci number.

Many have noticed how many times the price of gold has bounced off of its 144-day MA over the past few years (see blue dotted line on chart below). But, what's also remarkable is how linearly the moving average has increased since it began its upward trend in mid-February 2009. (Linear on a logarithmic chart, reflecting growth at 27% a year.)

I fit a regression line to the price of gold since early 2009 (not shown). The slope of that line has remained almost constant for years, and importantly, appears to have the exact same slope as that of the regression line of the 144-day MA (see brown line on chart below). Also note the grey dotted lines, representing ~1.25% in either direction of the 144-day MA's regression line: it's never gone beyond that.

This steady rate of increase of the moving average seems a little too perfect for me. Remember, this has been going on since gold was in the low $800's. It fits perfectly with what Bill Murphy of GATA has called a "managed retreat" by the banking cartel, and what Jim Rickards has explained is the government's wish with respect to gold: they want it to increase in price (i.e. they want the dollar to lose value) -- as long as it happens in a controlled manner.

Obviously, this trend won't go on forever. But what if it goes on for the rest of the year? Let's assume that the 144-day MA will indeed continue to trace the brown regression line on the chart above (or oscillate between the dotted grey lines, at any rate). We can then get a good prediction of the actual price of gold on any given day. That's because, with every passing day, we know which number will drop out of the 144-day moving average (e.g. tomorrow, gold's price on November 12, 2010 -- $1388.50 -- will drop out). Therefore, if the average is to continue increasing at 27%, we also know how much the gold price has to rise.

The x-axis of the chart below represents days since February 12, 2009. The y-axis marks the value of the above chart's brown regression line. If the 144-day MA continues its 27% rate of increase, then it should hit $1600 around November 21, 2011. If that's going to happen, then the price of gold has to average $1620 until then. If we imagine, for simplicity, that the price of gold increases at a linear rate forward from here, then it would have to go up about $1.40 a day, and reach $1700 on November 21, 2011, for the 144-day MA to be at $1600 on that date. It can of course fall $50 today and still end up averaging $1620 a day by November 21, but the more it falls in the short term, the higher than $1700 it would have to be by November 21 for the condition to hold.


Warren James said...


Warren James said...

(just kidding). GM, this is great stuff.

So in a nutshell, hitting the 144-day moving average - while ever the trend is in place - is a good 'buy' signal.

I think you captured another derivatives fingerprint. Awesome..

GM Jenkins said...

Thanks, Warren - the question is, how unusual is it for a commodity or stock to have a <150 day moving average rise at such a steady rate for such a long time? I quickly looked at a bunch of different ones, checking out various MA's, and nothing I saw seemed even remotely comparable.

Look at the past 7 months: the 144 d MA might as well be a straight line. Of course, a MA that moves in a straight line tells you what the price is going to be, as I explain above. Just look for what's dropping out of the MA, and make up for that to get back on the line...

Louis Cypher said...

I think you have found a magic number.

Warren James said...

Just my imagination, but for your chart it appears to dive (and touch) the 144MA every 6 months, suggesting it will do that again around end of July (at approx. $1475). That implies a market correction around that time but in fact if that price is reached then the primary trend is still intact. Cool, this and your Silver:10y treasury post paint a positive picture for the metals long term.

Anonymous said...

GM Jenkins,

thanks - very nice chart.

Does anyone of you know enough statistics to estimate how likely it is that this has happened by chance?


Anonymous said...

Victorthecleaner - By pure chance, it's at least one in a quadrillion :-) However, we don't necessarily need to assume dark forces are at hand. (They might be, but they're not essential to explain the formation).

For example, the 144-day MA is now pretty famous. So all the buyers of size know/think that this is a good place to buy. So whenever the price bumps against the 144, they go all in, as the risk-reward is favourable. It thus becomes a self-fulfilling prophecy. This lasts until it doesn't - i.e. an external event pushes the price below the 144, and it vanishes as a tool as quickly as it appeared.

The fact that it's a Fib number can be explained by the same reasoning. The algos' buy points have to be set somewhere. The fact that people have confidence in Fib numbers, makes them self-fulfilling prophecies too. So the Fibs get programmed into the algos for as long as they hold.

I'm a believer in TA, but only because I know other people believe in it. I have bought gold at the 144 on a couple of occasions - but only did so because I reckoned the chart would convince others to do the same. Pure game theory. TA has no value without others' faith in TA - period.

It's also why I'm so bearish on silver at the moment - TA's almost impossible, because all the old 'rules' got broken along with the parabola. Yeah, it may go above $40 over the summer, or it might hit $28. Anyone telling you it's definitely one or the other based on TA is either mad, self-deluded, or a lier.

Anonymous said...

Jeanne d'Arc,

concerning silver: It seems that LBMA backwardation is disappearing right now. So my guess is silver will outperform the other typical 'risk-on' trades. Let's wait and see if I'm right (I won't tell you the reasoning until I have sound data). This is explicitly not a recommendation to trade!


Robert LeRoy Parker said...


Do you have these silver bars for your database?

Claymore Trust

Anonymous said...


Interesting. I'd thought that backwardation was traditionally interpreted as a lack of confidence in future supply availability, so buyers would be prepared to pay more for immediate possession of the good (i.e. buying at spot price) than for potential delivery (i.e. buying futures). So I thought backwardation = higher spot prices, and contango = lower spot prices(Am quite happy for you to explain to me why I'm wrong about this!)

Regardless, my hunch is that silver will plough its channel between 33.5 and 39.0 for the summer. That's actually a pretty decent trading range, and normally I'd be all over it. But silver has this way of biting you in the arse...

Some of the silver miners are beginning to look jolly cheap, however. I think I'll be picking up a few of them over the summer. I mean, really? How low can they actually go? CDE at 23.78 - that's just crazy cheap.

Robert LeRoy Parker said...

Gold versus paper has a new excellent commentary on the state of the market.


Unfortunately he is moving to a pay based service which is extremely disappointing to me. But perhaps he will keep posting free pieces like this.

Warren James said...

I'm a paid subscriber to Adam's new service. It's a good write-up. I've always liked his stuff and I hope to learn to trade better.

P.S. Robert thanks for the link earlier, I'm adding that feed to the database, along with a few others.

GM Jenkins said...

TA has no value without others' faith in TA - period.

Well, game theory will make chart patterns self-fulfilling, but there's a lot of basic psychology too behind the fundamental ideas of TA, as I'm sure you know, such as when broken support becomes resistance: a price range is support bec a lot of people obviously bought there; then, when it gets broken and goes back up there, many are happy to unload and break even. I'm sure things like hammer formations on candlestick charts and outside reversals and stuff also have psychological grounding.

But on that note, and touching on Victor's question about whether this 144 thing can be statistically analyzed, it seems to me that while the 144-day MA as support can easily be attributed to algos and crowd psychology, the linear movement of the MA on the log chart (*very* linear over the past 7-8 months) is a lot harder to explain away that way. Who, when making an investment decision, looks at what happened to an asset 144 days ago, and decides to buy or sell on that basis? E.g., today, the gold price that dropped out of the 144-day MA was a relatively low price, right at the time of one of the steep November "corrections" (I believe - hard to be sure, bec I'm using raw data from the World Gold Council, which is at times slightly different than from stockcharts.com, which doesn't give raw data, and it's unlikely but possible I screwed up the nummber of trading days etc.). So, since a relatively small price dropped out 144 days ago, gold fell $10 today and still allowed the 144-day MA to climb up at close to the .09% level that it currently has to if it's going to hew close to its regression line since 2009 (the 144 MA climbed up .08% today). Despite this regularity, it's probably not possible to make accurate investment decision based on this pattern on a day-to-day basis, because the 144 MA could go up less than needed today and make up for it tomorrow, but then the further away you are, and the greater the price 144 days ago that's dropping out, the more confident you can be that gold needs to go up big today.

Re: statistically interpreting the likelihood of this occurring by chance (since you can, hopefully, assume a large crowd isn't using this "144 days apart" method to make their trading decisions) would be by looking at the correlations in % price movements 144 days apart over 2.5 yrs and compare that to correlations for 144-day spreads of as many different stocks as commodities as possible to be able to get an idea of how rare this event is. You'd need a lot of data and some programming skills.

GM Jenkins said...

Let me add that, if this linear-ish % rise of the 144 MA is to remain, gold should actually move up/sideways over the next 30 trading days, hitting 1575-1600, then, as Warren has noted, have a steep decline, corresponding to the January correction, near the end of July. That should last a month or so, but then it has to rise faster than it did in Feb 2011 (till now) to stay on the line.

It's funny how this is playing out to coincide with the end of QE2. I could see the correction happening at the beginning of July, i.e. three weeks early, pushing the 144-day MA towards the lower bound of the grey dotted lines, then with some announcement of further QE, really popping by summer's end to make up for it.

Or, as JdA has said, this pattern will hold until it doesn't and this is all bullshit speculation. But, personally, seeing as I believe this pattern is too perfect to be due to chance, I think I'll want to be out of gold by the end of the month, or, if it hits 1575-1600, even shorting it.

Swampfox said...


Thanks for the link. Interesting article.

Robert LeRoy Parker said...

Pretty crazy article.


Robert said...

Warren ,

you said you are a paid subscriber to Adam's new newsletter.Could you go into a little more detail if you think its worth it --How often does it come out,and maybe something about his trading advise--Any help would be appreciated .

Warren James said...

@Robert, the best advice I can suggest is to try it out for a month - cost is $15. The word:dollar ratio is favourable, he is currently sending material every 2-3 days and you get a deeper insight into what his thoughts are. My comment here is not an endorsement. His very first tip on ZSL got tripped up badly with the weird price movements in silver, but that's lesson #1 I suppose - that if one must use the stock market to skim fiat from the other players then you gotta do it without emotion and stick to basic strategies.

About half of what he writes is about the dow:gold ratio, which is why I subscribed. I guess you could say it's a detailed version of his blog content.

Given the price and the amount of time he must be putting into the write-up I am guessing that it's an exercise in sharpening his own understanding more than anything. HTH.

My own objectives are to just rake in a few extra dollars from trading to diversify my income streams, and perhaps build up some capital for a business venture. In this respect, GM's charts are useful because he's uncovered some nice clear patterns. In this distorted market, it seems that only the fundamentals will win out - it's my understanding that the wealth pump now created has been fine tuned to such a degree that it is impossible to escape from, to the point where even fundamentals go out the window. Sad really. My eye sees the whole show as a giant tumour/cancer growing on the face of humanity. It may not be terminal but it sure as hell causes a lot of needless suffering. I suppose that's why I hope that systemic, auto-balancing paradigms like Freegold will eventually manifest themselves.

Robert said...

Thanks for the reply--what the heck maybe give it a try for a month or so --sure can't hurt.There are so many out there that have paid services and only a handful at best are worth it.For my money one of the best is Bill Downey from Gold Trends .He does a nice job of combining TA with fundamentals .He has a great record and really knows the gold/silver markets.
Again thanks