Rethinking the WTI/gold ratio as oil fundamentals change

Maybe it's time to scrap these oil vs. gold comparisons. It's possible the historic relationship has collapsed into a new normal, and anybody trading an arbitrage expecting a return to the old norm might have a long, long wait.

We first took a year-end look at the spread between oil and gold several years ago, because they had long been seen as having a relationship that, over time, would drive them toward a standard ratio.

Both were widely traded, were seen as storehouses of value, and were often the place to run to when geopolitics got nasty. And their ratio of 15 — one ounce of gold could be exchanged for about 15 barrels of oil — seemed pretty consistent over time. If it was far from that, it could be viewed as a signal that something was going to have to give.

Note: We have always used WTI as the basis for the oil price, because the data goes back further than that of Brent. That was a problem for those years when the Brent/WTI spread blew out, but a more normal relationship between those two benchmarks now mostly negates that impact.

Consider: In 2014, the ratio of WTI/gold, until the very end of the year, was never wider than 2 percentage points relative to 15, on either side. With the exception of a few days, you can say the same for 2013. The average ratio for the first 20 years of our data — from 1984 through 2003 — was 17.34; that is, an ounce of gold fetched 17.34 barrels of oil. The average since then is 14.99, almost exactly at the conventional wisdom of what the ratio should be.

The ratio has averaged 18.08 since the day in July 2008 when oil hit an all-time high. The ratio that day, incidentally, was 6.43. The all-time low was 6.17 on August 30, 2005, when gold was all the way down to about $430/oz. and oil was just under $70.

In the last two years, it’s like a new world. The average ratio for 2016 was 29.54, following a 2015 ratio of 24.18. The 2016 average includes a period from the start of the year through early April when oil stayed steady around $35-$36 while gold surged from $1,080 per ounce to almost $1,220, boosting the ratio from 30.19 at the start of the year to 33.85. That included several days above 40, the first time in the history of our series we've seen that.

By the end of 2016, oil prices had risen off their sub-$30 lows, while gold — moving toward the norm of 15 — declined to less than $1,150/oz. That brought the ratio down to 21.58 by the end of the year, still 6.58 percentage points above the norm that maybe isn't a norm anymore. Perspective: If gold held at $1,150, oil would need to get to at least $77 before the ratio would return to 15.

Oil's fundamentals have changed for the foreseeable future because of the shale revolution. A possible analogy from the past: the shale revolution first hit natural gas, and its price plummeted relative to oil. A lot of traders figured the spread between the two would return to “normal” levels, and traded accordingly. They lost a lot of money. At the start of 2007, the price of a barrel of WTI was about 9 times the price of a thousand cubic feet of natural gas. It crossed double digits in the middle of that year, and  has not been single digits since. It was about 30 a few times in 2013, and averaged 17 in 2016.

Meanwhile, flight to safety and a concern about inflation, which at one time had both seemed to always involved going long gold, now is targeted at other assets; that can be seen in part in the surge of the dollar. A strong dollar is bearish for both oil and gold, but clearly not at the same rate. That's why an oil/gold ratio of 15 might be an old norm, and not returning soon.