This Gold Model Calculates Prices Between 1971 & 2017

Read the Latest News About:

Gold    Silver    Economy    Central Banking

Gary Christenson - Deviant Investor

Gold persistently rallied from 2001 to August 2011. Since then it has fallen rather hard – down nearly 40%. This begs the question:

What happens next?

  • Did the gold bull market end at the top in August 2011 as many mainstream analysts believe?
  • OR

  • Was the decline during the past 2.5 years merely a correction in the ongoing bull market?

The answer, in my opinion, can be found in my gold pricing model that has accurately replicated AVERAGE gold prices after the noise of politics, news, high frequency trading, and day-to-day “management” have been purged.

I presented the specifics of my model at the Liberty Mastermind Symposium in Las Vegas on February 22, 2014. A detailed presentation would be much too long for this article, so the following is a quick summary.

Like this blog? You might enjoy my e-book:

Survival Investing
with Gold & Silver

by GE Christenson – aka Deviant Investor

Kindle      Smashwords

Want to be an Affiliate?


  • Create a simple model of gold prices based on a few macro-economic variables, NOT including the price of gold.
  • Each variable must be intuitively sensible in its affect upon the price of gold.
  • The results must be graphically similar to actual prices for gold since 1971 and be statistically significant.


  • The most obvious macro-economic variable is the currency supply or some proxy for it. Since 1971 the U.S. currency supply has been increased much more rapidly than the underlying economy has grown. Hence the value (purchasing power) of each currency unit (dollars) decreased and prices, on average, have risen considerably.
  • Other variables that might be applicable are the CPI, Japanese Yen, real interest rates, dollar index, 30 year T-bond yields, DOW Index, copper prices, national debt, commodity prices, and many more.
  • A logical and causal relationship can be established between each of these variables and the value of gold based on either the declining value of the currency, or the changing demand for commodities and hard assets versus the demand for financial assets.


  • My model was created, tested, and refined to include only three variables – simplicity makes the model more credible.
  • My model attempted to replicate the smoothed annual prices for gold. Smoothing filtered out most of the market noise and clarified what I refer to as an equilibrium or “fair” value for gold.
  • My model made NO attempt to predict actual weekly and monthly gold market prices.
  • Smoothing was accomplished by using monthly closing prices for gold since 1971, creating a centered 13 month moving average of those prices, averaging January to December monthly prices to create an annual price, and then making a 3 year moving average of those annual prices.
  • Smoothing examples: Actual market prices in 1980 went as high as $850 but the smoothed gold price for 1980 was about $460. Actual market prices in December 2013 went to an approximate low of $1,183 but the smoothed gold price for 2013 was about $1,520.

Model Results

  • The calculated Equilibrium Gold Price (EGP) had a correlation of 0.98 with the smoothed gold price from 1971 – 2013. Examine the graph of EGP and Smoothed Gold.
  • The model was both simple and robust. It worked effectively, on average, during gold bull and bear markets, stock bull and bear markets, blow-off tops and crashes, volatile oil prices, Y2K and 9-11, QE, Operation Twist, ZIRP, various hot and cold wars, occasional peace, gold leasing, gold manipulations, and high frequency trading distortions in many markets.
  • In August of 2011 gold was priced about 30% ABOVE the EGP.
  • In December of 2013 gold was priced about 26% BELOW the EGP.

Graph Notes

  • Smoothed gold prices are shown in a “gold” color.
  • Calculated equilibrium gold prices (EGP) are shown in green.
  • The long-term trend from 1971 – 1980 was up, from 1980 – 2001 the trend was down, and from 2001 to 2012 the trend was up. (Actual gold market high price was August 2011.)
  • Nixon closed the “gold window” in 1971, removed any semblance of gold backing for the dollar, and thereby enabled the creation of significantly more dollars into circulation. The various measures of “money” supply, official national debt, Dow Index, price of gold, many commodities, and most other prices increased exponentially between 1971 and 2013.

Future Prices for Gold per the EGP Model


  • Macro-economic variables continue to increase and decrease as they have for the past 42 years.
  • The U.S. economy continues along its typical, but weakened, path with government expenses growing more rapidly than revenues, as they have for decades. National debt rises inevitably.
  • Congress continues its multi-decade habit of borrowing and spending, talking about change, and changing little. The Fed supports the stock and bond markets and continues “liquidity injections” as it deems appropriate to benefit the 1%.
  • Monetary, political, and fiscal policies will NOT be materially different from what they have been during the past 42 years.
  • The U.S. will NOT be subjected to global nuclear war, Weimar hyperinflation, or an economic collapse, while we will continue to be subjected to the same Keynesian economic nonsense that has created many of our current “challenges.”

Given the above assumptions, a reasonable projection for the EGP (a “fair” price for gold) in 2017 is $2,400 – $2,900. Remembering that market prices can spike significantly above or crash below the EGP for many months, we could see a spike high above $3,500 or $4,000 in 2017. Extraordinary events such as a global war or dollar melt-down could push prices higher and sooner.

I plan to publish the details of this model, including variables, graphs, analysis, and the calculation formula in a paperback book.

Until then, you may find value in these articles:

Bill Holter Jim Sinclair in Austin, Texas
Eric Sprott Do Western Central Banks Have Any Gold Left?
Gold Silver Worlds Jim Rickards: Target Gold Price
Casey Research 23 Reasons to Be Bullish on Gold
The DI Gold Investors: Take the Red Pill
Jim Sinclair MineSet

GE Christenson
aka Deviant Investor

If you would like to be updated on new blog posts, please subscribe to my RSS Feed or e-mail.

Promote, Share, or Save This Article
If you like this article, please consider bookmarking or helping us promote it!

10 thoughts on “This Gold Model Calculates Prices Between 1971 & 2017

  1. Hi, Just stumbled on your blog following its mention on 321 Gold. This is a superb article, well done. Also well done in your polite response to the comment from Vess. I would guess that most of us would explode with anger if our gold price model was attacked in such a cavalier manner, particularly when the attacker casually mentions that gold is heading to $1000 over the next decade. What arrogance!

  2. I read your article published in the Lemetropole Cafe on 3/4/14 and would like to know when you would be publishing the paperback book on the details of the model you had described.

    • Thank you for asking. The book is a “work in progress” and so I do not have a publication date. I have a hopeful target of July of this year.

      Thanks for your interest.

      GE Christenson
      aka Deviant Investor

  3. 1) With so much “smoothing” the EGP deviates so much from the actual gold price as to become meaningless. I doubt that anyone who has bought gold at $850 would be very happy with your model telling him that he bought at the average price of $460.

    2) Just eyeballing the graph suggests that during the gold secular bear market (1980-2000) the correlation between your model and the gold price was mostly negative.

    3) I have not seen any arguments that your model has predictive power. It didn’t tell you in 2001 that the price of gold will rise above $1000. Could it be that it similarly doesn’t tell you now that the price of gold will fall below $1000 within a decade?

    4) 2.5 years of the price trending down is not a “correction”. It is a full-blown bear market. Maybe it was in the context of a secular bull market – but we won’t know that until much further down the line. So, the real question is: has the cyclical bear market that we’ve had for the past 2.5 years ended or not? If it has, we’ll have at least a cyclical bull market, even if the secular bull market has ended. If not, we’ll have further downside even if the secular bull market hasn’t ended yet.

    • First of all, thank you for your comments.

      Secondly, I think, for the most part, you missed the point of the EGP model and the article. I’ll address your concerns individually.

      1) “With so much smoothing the EGP deviates so much from the actual gold price as to become meaningless…”

      This depends a great deal on what you think meaningless is. I did state that the model made NO attempt to predict weekly and monthly gold prices. So if you mean that the EQP is supposed to predict short term prices (which it isn’t) then, I agree, it has little short term predictive value. But what the EGP does is project an approximate value for gold – broadly speaking – at any particlular time.

      Regarding your comment that “I doubt anyone who bought gold at $850 etc….” Neither the model nor the article told that person he bought at an average price of $460. What the model did say was that a fair value for gold at the time was in the $400 – $500 range and not above $850. So, if you think it through, knowing that gold had a value of say $400 or $500, would it be sensible and prudent to buy gold at $850? Clearly not. HENCE THE VALUE OF THE MODEL!

      2) “Just eyeballing the graph….” The model made no effort to predict monthly prices. It did make an effort to identify long term trends which it clearly did in that secular bear market from 1980 – 2001. On a short term basis one line may have gone slightly upward at the moment another line was moving downward but EXCEL calculated the correlation between the two lines, over the 42 year period, at 0.98. Eyeballing is important but so are objective numbers such as the correlation provided by EXCEL.

      3) The model has, or does not have, predictive power, based on two issues:

      a) Is the price of gold over-valued (such as in August 2011 and Jan. 1980) or undervalued, (Dec. 2013)? If undervalued, anticipate that, on average, the price will go up. If overvalued, be careful. Prices also fall.

      b) The EGP is based on 3 variables (which I defined at the Liberty Mastermind Symposium but did not define in the article). If you can project those three variables going forward (not that difficult – in my opinion) then you have some predictive power going forward for the EGP.

      And no, the model does NOT tell me the price of gold will fall below $1000 in the next decade. In fact it encourages me to believe that $2400 – $2900 is reasonable for the EGP in 2017. It is a model, it is only a reasonable projection based on rational analysis and a few assumptions – it is not a prediction handed down from a higher source, but it is a great deal better than nothing, and a great deal better than self-serving forecasts put out by large trading firms who trade against their own public recommendations.

      4) “2.5 years of price trending down is not a correction. It is a full blown bear market.” Maybe. That is your opinion so I encourage you to short gold to stay true to your assessment of the market. However, I choose, in theory, to be on the other side of that trade as I do not see this as a bear market, but a correction in an on-going bull market.

      The model agrees that it is a continuation of a long term bull market, and is NOT similar to the 1980 – 2001 period.

      Another example: the T-Bond market has been in a bull market from roughly 1981 – until 2012, and possibly until the present. Have T-Bonds topped out or not? I don’t have an opinion as I claim no expertise in the bond market. If you thought that T-Bonds had topped out in 2012, you might be correct. But consider:

      T-Bonds peaked in October 1993 and dropped into lows in both April of 1997 and January of 2000. It could easily have been said that T-Bonds were in a bear market since 3.5 year later, in April of 1997, they were still far lower than their 1993 peak. In retrospect that was clearly incorrect.

      It is too soon, in my opinion, to declare this a bear market in gold. The EGP model strongly suggests it is not a bear market, and in fact, clearly states that the price of gold is currently undervalued and likely to rise into 2017.

      Good luck with your gold short.

      GE Christenson
      aka Deviant Investor

      • I am afraid I still don’t understand the point of your model.

        Does it have predictive power or not? If it does, how exactly did it predict that the price would go up several times in 2000? If it doesn’t, how can you claim to predict, based on it, what the price would be in 2017?

        At best, it could tell you if gold is undervalued or overvalued RIGHT NOW. That cannot predict which way the price will move, but it can help managing the risks. From your graph, the current EGP seems to be slightly above the smoothed gold price. The way I understand it, this means that, according to your model, the current smoothed gold price is slightly undervalued. That’s a reasonable observation.

        It doesn’t tell us where the price of gold would be in 2017, though, so I still don’t see how you came to your prediction. Furthermore, that same graph has plenty of examples where the model showed that the smoothed gold price was even more undervalued – yet the price kept going down. The way I see it, the current discrepancy between model and reality (as much as your smoothed gold price is reality, of course), is simply within the range of error of the model and doesn’t tell us anything useful.

        Yes, I got that over the a 42-year period your model shows a 0.98 correlation. But you have dodged my question. I was asking about the period covering ONLY the secular bear market in gold – 1980-2001. What was the correlation during this period ONLY? Wasn’t it NEGATIVE? It should be pretty easy to use Excel to determine that, no?

        I don’t know what variables your model is based on and how easy (and reliable) is to project them in the future. And, honestly, I don’t care. All I asked was about the predictive power of your model. Did it correctly predict the (smoothed) gold price that was reached a couple of years ago while gold was still in the depths of its secular bear? It either did or it did not. And if it didn’t, what exactly makes you confident to make predictions now based on that model?

        As I said, 2.5 years of the price trending down (and falling 30% from the top) is a full-blown bear market. We had a similar one in the 70s, smack in the middle of a secular bull market in gold. Has the secular bull market in gold ended in 2011? I don’t know. And I don’t care. Your model tells you that it hasn’t? Maybe it is wrong. Maybe not. I don’t care. All I care is whether the CYCLICAL BEAR IN GOLD that we’ve had for the past 2.5 years has ended or not.

        As I already explained, if it has ended, gold will go up REGARDLESS of whether the secular bull has ended or not. And if it hasn’t ended, gold will go down again REGARDLESS of whether the secular bull has ended or not. So, it is the question that really matters to me.

        I do not understand how your T-Bond example is relevant. I don’t know whether the secular bull there has ended, either. The fact that it had a cyclical bear 3.5 years long does not invalidate in any way my saying that gold has had a 2.5-year long bear market. If you want to underscore that the 2.5-year bear market in gold does not necessarily mean that the secular bull market in gold that began in 2001 is over – I can agree with that. It’s just that it is irrelevant to me, as I have explained above. I am just objecting against your calling a 2.5-year long bear market a “correction”. At best, it is a cyclical bear in the context of a secular bull.

        I don’t know how you decided that I am shorting gold. In fact, I don’t short monetary metals as a matter of principle.

        My own analysis (using conventional technical analysis, not weird models) suggests that the cyclical bear in gold has, indeed, ended in December. I sold my holdings about a month before it began; currently I am inching in bit by bit. Not betting the farm, of course, because I could be wrong and we could still see new lows.

        • In my opinion:

          1) Yes, it has predictive power, but it is only a model, so it obviously has limitations.

          2) Yes, it does say that at times, such as now, actual gold market prices are lower than the EGP. Hence the strong inclination is that gold prices should move upward. Similarly in August of 2011 gold market prices were, according to the model, too high. Hence the model indicated that prices should come down – as in highly likely but obviously not guaranteed.

          3) Correlation over 42 years was 0.98. I did not dodge question about the correlation between 1980 and 2001 – I simply ignored it as largely irrelevant. However, since you asked, Excel calculated the correlation as +0.56 over that 21 year period. Graphically the prices of the EGP and the Smoothed gold price were generally parallel over the long term, but often moved slightly away from each other in the short term so 0.56 makes sense.

          4) re Predictive Power: I did some very rough back testing. Thanks for your insistent prodding. This back testing needs more much work and more time, but in the interests of getting an answer to you, I did the following.

          Assume 3 year future projection is feasible and long enough to be relevant. So, during 2001, what did the model indicate for 2004? Similarly, from 2003 what did it indicate for 2006? To be as intellectually honest as possible, I had to make assumptions based on the history of the variables – not the current day events. Remember, much changed on 9/11/2001 and the calculation from 2001 did not know about 9-11, and the calculation from 2003 did not make any allowances for 9-11 – in my effort to be intellectually honest and consistent.

          Conclusion: 2001 average smoothed price was about $275, and 2004 EGP projection was about $375 while the smoothed price for 2004 was about $410.

          Conclusion: 2003 average smoothed price was about $365, and 2006 EGP projection was about $505 while the smoothed price for 2006 was about $605.

          Given that 9-11 changed our fiscal and monetary world in so many ways, and the above calculations took no account for those changes, I would say (and you might disagree) that the model had predictive power – even at a turning point in gold prices such as the 2001 – 2002 time period.

          5) As for your comments about cyclic bear markets etc, I’m not going down that rabbit hole again. Call it whatever name you wish, but the issue is: Will the price of gold be much higher, about the same, or lower in 3 years. The EGP model indicates much higher.

          6) Congratulations to you if you sold out near the top and watched the market move downward from the sidelines. I expect you are one of the few.

          7) Let me know when and why you decide to go back long the gold market.

          GE Christenson
          aka Deviant Investor

          • It seems that we are speaking different languages, so I won’t bother with this discussion any longer.

            Regarding point 7 – I already gave you an answer in my previous message. I have started taking positions since January, because my technical analysis told me that the cyclical bear has most likely ended. If you are asking when I’ll be “all in”, the answer is: NEVER. I will never bet all my life savings on any one thing – not gold, not anything else. I am just trying to take positions in assets that are appreciating and get out of them when my analysis suggests that their price is overstretched.

            Where will gold be in 3 years? Probably higher than today. Much, much higher? Most probably not. Much lower? Again, most probably not. But I will react (and will get out of it) when my analysis suggests that a top is approaching. If it gets in a bubble (and that’s a big if), that will mostly likely be much before the actual top, price-wise. Being just a month early in 1980 meant leaving a lot on the table. But that’s preferable to me that taking unnecessary risks.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.