A curiosity: Over the period from Friday, September 4, through Thursday, September 10, the price of UNG units in the market ran up from the close on September 3 of $9.01 to a close on the 10th (the Thursday of the weekly EIA NG reports) of $11.18, a 24.08% increase.
This was accompanied by a puzzling run-up in NG prices. (Click chart to enlarge.)
The reason I say "puzzling" is that the weekly EIA summary report for the previous week, issued 9/3, indicated gas in storage still 18% above 5 year averages and reported a net injection of 65 Bcf, slightly higher than the 5 year average of 64 Bcf but below last year's 92 Bcf by 29%. The way net injections have trended over the last five or six weeks, I think there is absolutely nothing unexpected or unusual about this net injection. Storage was 17.8% above the 5 year average, a slight drop from the highs in the previous weeks. Still, nothing to cause one to turn hand-springs over.
In the EIA NG Weekly Update, this week's injection of 69 Bcf was slightly higher than the 5 year average of 67 Bcf and last years injection of 63 Bcf. This leaves storage at 17.1% over last year and 17.4% over the five year average. This is a mild improvement over previous weeks "overstock", but should have been expected since rigs were dropping like flies until 6 or 7 weeks ago. In spite of this, NG finished higher on Thursday.
I heard a lot of chatter on CNBC saying the traders said short covering caused the price rise. Maybe this explains it. Maybe not. We'll consider this more a little later.
A new gas in storage record is expected at the end of injection season - 3,842 Bcf. Winter weather is predicted milder than normal, hurricane season has been a real dud, Baker-Hughes rig counts increased 7 weeks straight, demand is an anachronism that used to describe what justified producing a product, EIA expects October prices to average $2.25, everybody is expecting a big ramp-up in LNG imports, and the latest EIA-reported consumption (May and June) is in line with previous year's norms.
Nevertheless, CNBC reported the traders viewed this injection as "less than expected". Huh?!
Is it possible the traders figured all the new rigs over the last 7 weeks - now back up to 699 (2003 levels and the previous week it was 701) from 665 on 7/17, +5% - got new wells operational, hooked up to pipelines and pressurizing already? Did they expect a veritable tsunami of new injections would occur? Does this explain the view that it was "less than expected".
I'm sure the traders have at least as much information as I. And they certainly have much more experience. Combine the two and you know they are not going to be too far off the estimates of net injections. If this injection is within what I would expect (and it was), give or take a couple percent, you know they had it nailed.
I really would like to know what caused this price behavior and why the latest injection was "less than expected". As part of my steep learning curve I'm not afraid to ask questions and look ignorant. I just don't want to remain ignorant.
Anyone care to fill me in? I certainly would appreciate it.
Until I learn more, I don't believe I heard the truth.
What about the "short covering"?
The September contract closed August 27 and the October contract closes September 28. If you were short, would you be covering during a 25%-30% price increase knowing all the above? Would you be causing it? With two full weeks left until trading termination and being aware of all the above, would you feel pressured (or fearful) enough to cover now as prices ran up 25%-30%?
So, either there is something I have overlooked, certainly possible and even likely, something that I'm still really ignorant about (lots of those things, still) or there is something else going on.
Could options on NG derivatives cause this? I don't know enough about how those work to say. But even those options can trade through September 17. If everybody who was long a put closed out their positions to take profit, could such a run up be caused? What if they exercised them? Would this cause such price behavior? Certainly folks know that NG is heading down and would wait for better profit on those puts, wouldn't they? Like I said, I'm just too ignorant of this stuff to know.
What about long calls? Could a massive exercise of those cause it? This seems a distinct possibility. But if the holders knew the above, why would they hold calls now? Is it possible that some call writers were naked, got assigned and had to cover? Again, "I plead the fifth" since I'm ignorant of the mechanics of these things.
Any help on the mechanics of this would be helpful and appreciated.
On September 11, NG reversed course, getting as low as $2.77, if I recall. Last settle at 17:15 on the October contract was $2.96 and last trade was $2.98, -$0.296. I feel much better now. It's at least moving as I would expect, given my minimal knowledge and experience. If it didn't do this at some point, I would have to toss all I had worked so hard to learn as being erroneous. There's still so much to learn, but I at least felt I had begun to acquire a foundation.
Anyway, on with the story. Lets take a quick look at the UNG 200 day chart and note the nice, calm, as-expected downward trend over the last few weeks or so. Then notice what may have terrorized shorts in UNG the last three or four days. (Click to enlarge.)
Let's talk about what UNG did on Friday. UNG nose-dived back down to as low as $10.50 and closed at $10.59. If the previous ratio of the underlying futures contract to UNG NAV is still 3.08:1 (marked down from 3.45:1 formerly) and NG October futures contract closed at $2.98, UNG NAV should be around $9.18. The UNG web site shows $9.12, so I'm close enough for government work. Add in the premium. So lets use UNG site's figures $9.12 + 16.12% = $10.59. Ain't math "wunnerful"? Exactly what we'll see in the chart below.
This dive was accompanied by a couple of unusual "events".
As will be noted on the 1 minute chart of UNG trading for the day, a large price drop on very high volume began at 12:19 and ran through 14:00. Volume for this run was 19.3MM trades. This was 29.8% of the days total volume of 64.7MM trades in a time span of 1:45. Having watched UNG trading many months, I noticed these "events" and the "waterfall pattern" immediately as they were strong variations from the normal observed daily minute-by-minute price action profiles. Even allowing for the volatility of the spot NG price today, which varied up and down at various times in a range of approximately $2.93 to $3.45, this is highly unusual. (Click chart to enlarge.)
The few breaks in the down trend were all short and weak and had little breadth, another unusual variation.
Given the volatility of NG today, one might think that UNG could track it pretty closely and be just as volatile. That may be. But I have not observed this type of behavior over several months on days when NG was more volatile than usual. UNG trading tends to follow some general patterns of price swings throughout the day, less sensitive to the minute-by-minute futures or spot price changes of NG and tends to take a final leg down or up to "gravitate" towards the NG-dictated price approaching the end of trading. Seldom are big persistent uni-directional moves seen in the middle of the day. Most moves from late morning through early afternoon tend to have common patterns of up, down, consolidate a bit, repeat. And there will be some runs up $0.20 or so and similar on the way down. But overall, you see trends, breaks, reversals and consolidations over time.
On quiet days, the price range encompasses $0.20-$0.40 or so. On more volatile days, usually swings around $0.70 or so. Both ranges are rough estimates from observation and are not calculated. Regardless, the patterns tend to be as I described above.
The range for 9/11 was not highly unusual - $10.50 to $11.37. But the "waterfall" runs downward were quite exceptional. I can't recall ever seeing this pattern occur so sharply and multiple times on a continuous extended down trend in the middle of the day like this, from $11.24 down to $10.50, a 6.6% decline.
That certainly caught my eye. Then something happened.
The new UNG 8-K was listed as EFFECTIVE 2009-09-11 17:02:47. A check of the directory timestamps and the parent directory timestamps are consistent, indicating a high likelihood that these timestamps are accurate and uncorrupted. A snapshot of these items can be seen in the next three graphics. (Click graphics to enlarge.)
My first news notification, via my Power E*Trade Pro trading platform was timestamped at 17:06:16 from the Dow Jones Newswires service, as shown in the following screen-shot. The full article of the announcement was timestamped 17:38:28. (Click to enlarge ... I think.)
Well, the old "associative processor" kicked right in.
Was this related to the unusual price action I observed in UNG earlier? [Queue the theme from Jaws Maestro.]
Without access to minute-by-minute NG October futures prices, I have no way to "triage" my thoughts and identify most likely scenarios first. Given all I stated above, you already know that I believe the rise in the NG price was an oddity and had no discernible reason to occur.
That's all I needed.
Was there some manipulation and/or "front running" here? My thoughts followed this path.
- UNG has been working, for an extended time, on a structure and arrangements so that it could begin issuing creation baskets again.
- In this process, there were likely discussions needed with outside parties, possibly consultants, "authorized purchasers", exchange authorities, financial parties, their brokering party, etc.
- Plans would need to be developed and firmed up and, one could assume, coordination with multiple parties would be required to make sure all the ducks were in a row.
- "Loose lips sink ships" and may also leak useful information to interested parties.
- Someone with knowledge of UNG's implementation plans and the wherewithal would be in a position to benefit if they could cause a rise in the price of UNG prior to the announcement.
- The most certain method to do this would be to cause a rising trend over several days in the October futures contracts, using them directly or possibly by working through the spot price.
- This might also be accomplished, regardless of wherewithal, by influencing traders expectations of net injections or other factors indicating a reason that NG prices should be expected to go higher.
- Selling UNG short into the resulting rise, in unobtrusive amounts over several days, would leave one well positioned for the coming downturn. Also if one were long a ton of UNG, this could provide a dumping ground.
- With the enthusiasm for all things NG, sell siders observing the NG/UNG rise would probably get their clients to provide the needed liquidity; they could be completely ignorant of what is going on. But just look at the volume over the last five days. Another nice ramp-up low volume was 46.8MM trades and high was 64.7MM trades in those last five days.
I'd say that was liquid, no?
Well, that's it in a nutshell. Have I gone over the edge?
Should I contact the SEC or CFTC? Or somebody effective?
How about someone's mother-in-law. That would likely be more useful.
I'm not one to lean towards "conspiracy theories" often, but this was just such a golden opportunity.
I look forward to any reasoned critiques, answers that help me along my learning path and other reasonable comments.
Disclosure: long UNG (before problems started), short calls, long puts, long synthetic shorts.