Lessons From a Failed Prediction
Why Make Predictions?
While there are actually other stories unfolding in the world of energy, you would never know that by my inbox. Most of the correspondence I have received in the past week is still related to oil prices, particularly following the recent huge rally in crude futures. A few readers also wanted to make sure that I noticed that one of my 2015 predictions had fallen last week. I will address that in today’s column.
For background, each year in January I make predictions for the upcoming year, and I provide the context for those predictions. (See My 2015 Energy Predictions). I have been doing this for several years, and at the end of each year I grade my predictions. As I have stated on many occasions, context around a prediction can be more important than the prediction itself. When I grade the predictions, I will talk about the context when I made each prediction, and the reasons the predictions turned out to be right or wrong.
But one reader asked why I would even attempt to make predictions given such uncertain conditions. I make predictions to set up a narrative that describes what I see unfolding in the energy sector, incorporating as much data as I can into making each prediction. While this is not an investment column, I am aware that some readers use it for investment advice. So without overtly recommending investments, I generally try to make predictions that are actionable. I will give 2 examples of that today, one of which is the prediction that failed last week.
Most of the predictions I make will require the full year’s worth of data to determine whether they are accurate. For example, one of my predictions this year was “The average Henry Hub spot price for natural gas will be below $3.50/MMBtu in 2015.” While that prediction is on track to be correct, we will need to be at or near the end of the year to be certain. A big price swing between now and the end of the year could change the result. This was the case with my 2014 prediction that the average closing price of WTI would be lower than in 2013. That prediction was trending incorrect until the 2nd half of the year when the price dropped so sharply.
An Oil Price Prediction Falls
I made two 2015 predictions, however, that could be proven right or wrong during the course of the year. One of those predictions was “The closing price of West Texas Intermediate (WTI) crude will not fall below $40/bbl in 2015.” As we know, WTI did close below $40/bbl last week, making this the first of my 2015 predictions to fail.
I began to get emails from people almost immediately after it happened. Some of them gloated “Ha! You aren’t so smart now, are you?” It was mostly the crowd predicting $20 oil who claimed that this proves that they know more about the oil market than I do. Judging from the action in the oil market the past 4 sessions, they may have gloated too soon. If they were betting on further price drops they may have suffered steep losses as the price of oil swung hard in the opposite direction.
Let’s address first things first. Yes, that prediction was wrong. When I tally up the score at the end of the year, that one goes into the “Wrong” bucket. In fact, at this point it looks like at least 3 of my 6 predictions will be wrong because they were strongly dependent upon the price of oil. Unless the price of WTI spends the rest of the year significantly higher than it is now, 3 predictions will fall on that basis.
If you care about the prediction more than the context, then you can stop reading right here and say “The prediction was wrong. End of story.” If you care about context, then what follows is an explanation of why I believe the narrative I used to support this prediction still has value.
A Useful Narrative?
A prediction can be wrong, and the context still correct and actionable. Why did I choose $40 as my line in the sand? The exact number was arbitrary. I tried to factor in supply, demand, and investor psychology. The higher I made the floor, the more aggressive the prediction. But the context for the prediction was “In today’s world, $40 is so ridiculously cheap for crude that I don’t think we will get there.” When I made the prediction, I wrote:
Of all the predictions I am making, this is the one with the greatest potential for being proven wrong the quickest. We aren’t all that far from $40/bbl today, and we could get there with 3 or 4 down sessions in a row. But if oil does drop below $40/bbl, it won’t be there long.
I considered this my riskiest prediction when I made it, because WTI was already in the $40s and dropping at a rate that would have taken it below $40 by the end of January unless it changed course. So it was a prediction that had the potential to be wrong shortly after I made it. Nevertheless, I still made it.
But the prediction survived until nearly September. On January 28th WTI closed at $44.08, but then ran back up to $53.56 over the following 3 weeks. Then WTI retraced to $44.02 by March 19th on concerns of growing crude oil inventories before running back up and bouncing around $60/bbl for most of May and June. So for most of the year, the price had proved to be pretty resistant to $40 — even though each time the price reached the low $40s, there were predictions that oil would continue to fall, possibly to as low as $10/bbl. It was only last week that the price finally closed below $40, and it only spent 2 sessions below $40 before the price surged.
What action might someone have taken on the basis of this prediction? If they believed the prediction, they might have viewed $40 as a buying opportunity for oil futures or oil companies. But if I called a floor at $40, and the price actually fell to $38.24 (the closing price of WTI on 8/24/15), was that a useful prediction? It all depends on what happens afterward. As I write these words, the price of WTI is now back up to $49.20 after surging over 28% since the lows of a week ago. Beleaguered oil companies have notched double-digit gains across the board. If you bought at $40 then you are happy right now, even though you might have bought just a bit lower. If the price plummets again and falls to $30 or lower, then in addition to the prediction of a $40 floor being wrong, the context will have been wrong and it will have cost investors (including myself) real money.
Those who are writing to me gloating are those who believe the narrative is wrong. They are the ones who believe that oil prices will go down much further. I still don’t see that happening. So whether the bottom was $40 or whether it was $38.24, I still view this as an unsustainably low price for oil in today’s world. U.S. oil production is now falling in response to low prices (and the EIA just revised U.S. production down again), and global demand is increasing by about 150,000 bpd per month. On the other hand, crude oil inventories are still high. The next major price move by oil will probably be dictated by how quickly falling production and rising demand can put a dent in crude oil inventories.
Checking in on the BP Prediction
By the way, I said above that 2 of my 2015 predictions could be proven right or wrong before year end. The prediction on a $40 floor for WTI was proven wrong. The other — “BP will be bought out or merged in 2015” — only has the potential to be proven correct before year end.
Again, there is the prediction, and there is the narrative that was behind the prediction. BP being bought out in 2015 was an aggressive prediction for several reasons, and while the $40 floor price for WTI was the riskiest prediction, as I noted when I made it, the BP prediction was “my most aggressive, wild card prediction for 2015.” A lot of things admittedly had to fall into place for that to happen in 2015, but the narrative behind this prediction was that because of their huge, high-profile missteps, BP would be more valuable under a different label. BP knows this, and their competitors know it. When I made the prediction BP’s proved reserves were on the books for a much lower value than those of their competitors. A person who accepted that reasoning might have opted to invest in BP instead of one of the other oil supermajors.
In fact, I have regularly argued this year in my weekly columns at Investing Daily that BP is grossly undervalued relative to the other oil supermajors. And even though it has been a bad market all around the energy sector this year, BP has outperformed its other supermajor competitors. Year-to-date, the share price of BP has performed 6.1% better than ExxonMobil (NYSE: XOM), 8.7% better than Shell (NYSE: RDS-A), and 15.6% better than Chevron (NYSE: CVX). So this could ultimately turn out like the $40 prediction — wrong, but with a useful (and actionable) narrative.
Wrong Prediction, Wrong Narrative
Now this doesn’t mean all the predictions I made this year had narratives that were ultimately of value. I have at least 2 where the narrative hasn’t played out like I thought it would, because oil prices were unable to hold onto the $60/bbl level they reached back in May. So when I am grading my predictions, if $38.24 holds as the floor I will consider that not too bad. I will grade myself much harsher on my predictions for the average oil price in 2015, and the performance of the Energy Select Sector SPDR ETF (XLE). As it stands right now — and as will probably be the case at year-end — these will be situations where the prediction and the narrative were both wrong.
Some of you may know that I write for a number of outlets. I employ different styles for different outlets. Here, I am more interested in what’s going on in the energy world without as much focus on the investment angle. When I write for Investing Daily, I will dig deeper into data supporting or undermining an investment case. I have done that for both the $40/bbl oil prediction, and the BP merger prediction.
Following my January predictions, I had a reader complain “You aren’t analyzing enough data to reach your conclusions.” It is true that I didn’t reproduce all the data here, but that doesn’t mean it wasn’t produced and analyzed. I am not flipping a coin to make these predictions. But the prediction column isn’t meant to be one that encompasses all the data. The six predictions I made could have each supported multiple columns (and several have).
I could address these concerns by frequently linking to Investing Daily articles that analyze the data in more depth. I have written at least 3 columns on BP, and probably 5 this year on oil prices for Investing Daily. But I don’t link to them much for two reasons. One is that you often have to wade through ads to get to the column. Regular readers here may be put off by that (and I don’t control it). Second, some of the articles end up behind a paywall anyway. So readers that are accustomed to one experience here will get a different feel over there.
Once more, because I won’t be able to say it enough to satisfy some critics, I predicted the price of WTI would not close below $40/bbl this year. It did last week, so that prediction fails. But as I advised in last week’s Investing Daily column — Panic Creates Opportunity for those who are well-informed and unafraid of short-term losses. If you are a long-term investor, $40/bbl oil is an opportunity that hasn’t been seen since 2008-9, after which the oil markets went on a five-year bull run. That’s unlikely to happen again, but I think energy stocks should outperform the broader market from $40/bbl. Or, in my opinion the upside at that point is greater than the downside.
If you are an investor who agrees with the rationale behind my $40 WTI prediction, then you likely agree that the narrative is more important than whether oil bottomed at $40 or at $38. Prices are just shy of $50 as I write this, but the operative word here continues to be volatility. The oil markets can change quickly. What looks like a buy today may not look like that in 3 months. The path forward is likely to continue to be rocky, but if you have the courage of your convictions I believe dips like we saw last week should be viewed as opportunities for long-term investors given the world’s current supply/demand outlook.
Link to Original Article: Lessons From a Failed Prediction