Why do gas prices show these strange patterns?

I am a member of Gasbuddy, so, while I could ask this on their forum, I’ve found that’s it’s a pretty pointless place for in-depth discussions, so I’m asking here.

Anyhoo, first of all go and click on this chart. The settings are NOT saved in the link, so you’ll need to manually customize the parameters:

1 Year
USA Average
Florida Average [I live in Jacksonville, and you can click on said city instead, won’t change much]

So, questions:

#1: Why are increases often so abrupt and large, while decreases so small and incremental? We’ve had two ~6-10 cent increases in the last 2 weeks, all happening on one day*, then statis for a week, then another one day multi-cent rise. But earlier in the summer, I’d note 1-2 cent drops once every day or three at just about every local station. The prices the stations pay to the companies who drive the tankers would presumably change at the same rate no matter if there was an increase or a decrease. So why the huge once-a-week rises vs. the small decreases several times a week?

#2: When prices are dropping, note how Florida’s average tends to fall quite a bit below the national one. But during these increases it will often equal or exceed the national mean. In these last two weeks the national went up only 6 cents, while Florida’s increased by 14. Why the discrepancy?

#3: Why, during these increases, are the stations often in total lockstep? If I had a dollar for every station which had their regular @ $2.28 today, I’d be a rather rich man now. Yet, when they are falling they will often get pretty far out of whack with each other (and this most notably applies to premium, which is what my car uses, with some stations for the same brand selling premium at 20-30 cents cheaper than others).

[*The apparent several day increases seen on the chart are almost certainly due to the Gasbuddy volunteers not instantly noting the changes until several or even many hours after the fact.]

It is a bit soon for wild-ass guessing, but I’ll take a stab at it.

What I think you are seeing is the effect of “cost of replacement” driving the cost at the pump. That is, when the refinery raises its prices, that communication is nearly instantaneous. Retailers raise their prices immediately so they can replace their inventory with the higher-priced product. If they don’t, they aren’t likely to have enough money to pay for the higher-priced product.

When price from the refinery drops, however, the retailer needs to get his return on the cost of the product he has in his inventory. If he dropped his prices immediately, he might not get as much for his inventory as he paid for it.

Of course, there are competitive forces acting on the system, which means there is a limit to how long the retailer can still get the higher prices for the goods. The retailer may have to sell some of his inventory at the lower price even though he paid a higher price for it because his competitors have lowered their prices and that is driving away customers. This just reinforces the need to raise the price immediately when they find out the refinery is raising the price, even though their tanks may be full at the lower cost.

Remember, the retailer has to make money. He cannot lose money or just break even. If he does not make money, he goes out of business.

#1 : The reason gas station prices rise and fall by different rules has to do with information flow.

It’s pretty simple. When the prices from the *supplier *for the gasoline the station is selling rise, the station owners find out within hours - they look it up online or they make a phone call.

The station owner isn’t in the business of losing money, so they immediately raise their price to the price from the supplier + a few cents per gallon in profit. That’s why you see those signs at exactly $2.28 or whatever.

The reason price decreases are much slower is the owner doesn’t care now what the supply price is. The owner is trying to extract as much profit as possible. So as long as a reasonable volume of cars are coming to the station, they leave the prices high even though the price for the supply of gasoline has decreased.

As sales volume starts to decline, they lower their price in increments, and so on. This is why gas stations on well trafficked streets seem to lower their price more slowly.

The difference in speed is fundamentally :

   a.  Drivers are slow to figure out which of the nearby gas stations offer the cheapest price
   b.  Station owners have a delay before they find out what their competitor stations have lowered their price to.  They will not lower their price until they are essentially forced to, so the prices stay high longer.

Change the location to Houston, Texas. There, the retailers are much closer to the refineries, so they aren’t as likely to be dealing with distributors that have large inventories they need to recoup. Same as I said above, but with the focus on the distributor rather than the retailer.

I used to fill up at an independant place that had a very large storage capacity and only supplied diesel to trucks. The owner was buying on the open market and having the storage allowed him to ride out peaks and take full advantage of troughs. So long as the average price kept increasing he made a good profit.

This worked out fine until 2009 when the price crashed. He had too much fuel stored and then he had to sell it at a loss because all his competitors had reduced their prices. AFAIK the land now has a shopping mall on it.

DesertRoomie worked for a time as assistant manager at a chain gas station. One of her duties was to report to corporate the prices of the three other stations that were within about 3/4-mile from hers by 7am; she generally did it off the clock as she was driving in. A couple hours later, a message would come back from HQ what their prices were to be for the day. She had no idea what algorithm (if any) they used but the price was always within a few cents - sometimes higher, sometimes lower – of the competitors, usually about a nickel above cost.

Then somebody dug through the 6-inch pipeline bringing gasoline from Texas to Tucson and Phoenix metro, the two largest population centers in the state. Within hours corporate started trucking gas up from Texas and the cost shot up to $5 per gallon, plus local delivery. The same procedure for pricing was followed and during the crisis it settled at about $4.50 a gallon. People screamed but her station, at least, money was lost on every sale; it had to be made up with cigarettes and jerky. Presumably the competitors were in the same boat.

After the pipeline was repaired, delivered costs went back to normal but the prices drifted down over a two week period as everyone recouped their losses.

@bob++: A hell of a lot of fuel-intensive businesses played that game. Which was a pure money spinner … until it wasn’t.

Google [fuel price hedge] for more.

Southwest Airlines, anyone? :rolleyes:

SWA used to brag that they ran a profitable airline every single quarter throughout the 1990s & 2000s when nobody else could make a profit hardly ever.

Not quite.

They actually ran a profitable commodity speculation business hampered by being attached to a loss-making airline most quarters. And once the speculation went South, they’ve been like the rest of us: profitable when fuel is cheap *and *the consumer economy is good, and unprofitable when either or both are otherwise.

If you don’t mind a hijack, how accurate do you folk find Gasbuddy?

My new car appreciates premium, but W of Chicago, the premium prices reported on Gasbuddy seem to have no relation to what I find when I pull up to the pump.

I pretty much agree with that and would just add that the sudden increase when the wholesale price is rising, to whatever it will cost to refill the tanks, regardless of what was paid for the gas now in the tanks, is what makes sense in an efficient market.

The part that defies efficiency is where the stations don’t immediately cut their price to reflect what it will cost to refill the tanks if that’s lower than what they paid for the gas now in the tanks. Of course they don’t want to do that, but in an efficient market they’d be forced to. In the professional market they’d have to: no other trader cares what you paid for your current position, just the price now.

The reason the station owners can get away with following their instinct and looking in the rear view mirror to what they paid for the gas now in the tanks, when the price is falling, is that consumers don’t immediately shift to the lowest price station, as professional traders would immediately shift to the lowest priced provider.

Partly that’s information asymmetry. But partly people who don’t really care that much what they pay for gas, despite the tendency of some others to over-obsess about gas prices (pay $2 more in gas burned to get to a far away cheap station to save 1, etc) and the general cultural norm (in the US certainly) of complaining about gas prices whenever they aren't really low. There's a station on a road I drive that's consistently more expensive than average, enough to be several 's a fill up, and the consistently cheapest one is only around 1 mile away though not along the same road. Both national brands, and one is not that much more jammed with cars than the other. This has persisted for years. So one element of the market environment is short term information asymmetry affecting most/all gasoline consumers, but another is persistent price insensitivity by many gasoline consumers.

Gasbuddy etc in theory would address the first issue of information. But it’s a pretty limited tool IME (I use it on road trips). For one thing (since at least last I used it) it doesn’t distinguish cash/credit prices which with a 5% back credit card is significant. And a lot of areas are pretty sparsely covered with up to date info. But again if the issue were purely info flow that would be one thing, but the other part is people who don’t care that much (and perhaps reasonably if it’s a few bucks a tank) what they pay for gas, really. They don’t use let alone contribute reports to GB.

As always xkcd: Working is on point. Randall teaches us that folks choosing to ignore local price variations in favor of habit or convenience are usually the smarter ones.

So the point is not that the consumers don’t *care *about small prices differences they know of. It’s just that in the physical world, unlike in the nearly frictionless trading world, there are real costs to switching providers. So the price delta needs to be bigger than the switching cost increment before there’s an arbitragable profit to switching.

The behavior may not be 100.000% rational. But it’s rational enough.

Welp, that’s a different topic, but, yeah. You see, they have this cute little preschool points reward system for reporting prices. But, of course, all that does is attract people who game the system to rack up the points (ostensibly for a chance at their adorable little lottery prize system), with the result that many people “reporting” prices have never actually visited/driven by that specific station that day. And yes for Premium it is much much worse, since many stations do not display the premium prices on their marquees.

Oh, you could just leave the non-displayed prices unreported (as your ghost goes zipping by the station), since there is an option to do exactly that, but, alas, no extra points for you. I just checked, they have one local station at $2.53 for premium–a mere one cent higher than plus. It almost certainly is actually 26 cents higher.

Randall apparently never has lived in Orlando.

  1. The economic adage is prices go up like a rocket and down like a feather. This is because no one asks for a second opinion when the news is good. If you go to the gas station and the price is higher than you were anticipating you are likely to shop around, and if the price is lower than your were anticipating you are less likely to shop around. This is less true for markets that have a Speedway gas station presence. They are more aggressive about cutting prices and so prices fall faster in those areas.
  2. State laws often create unique blends of gasoline that can only be used in that state or a state with similar laws. Usually demand for these blends mirror national prices but not always. If there is a weather event or something happens at a refinery that process a certain blend, prices for a state blend can get out of whack.
  3. Same answer as one. However some stations are not affiliated with chains and buy their gas on the spot market. Chains buy gas in advance and occasionally the spot market will either go significantly above or below the expected price.

Orlando sure is “special” that way. As I wrote about here in another context: http://boards.straightdope.com/sdmb/showthread.php?p=19756967#post19756967

Recoup their loses ? There’s an effect. To avoid the month or 3 month bottom line looking bad, they actually take a conservative approach to pricing, and sell less fuel, but make a profit.

If they start engaging in price cutting, then they start gaining market share, but not making a profit . So they would only engage in price cutting at a time their important goals will be met , eg at the end of month ? at the end of the three month period?

But well all the managers in the area want to avoid bad bottom lines showing up on their record, so they all avoid cut throat pricing in a crisis period…

For years we had a 7 day cycle here: Sharp rise very late Tuesday night, drift down to a low again next Tuesday. I don’t think that was driven by delivery prices.

Whatever it’s origin, at the end it was partly driven by the big fuel companies: they used it to drive the small independents out of business, by capturing the huge Tuesday evening market with big forecourts and lots of pumps. And it was eventually killed by the loss of the small independents, and the effect of Supermarket fuel coupons, which drove a lot of business to Saturday nights,

It was self-perpetuating for a long time: everybody who was price sensitive bought on Tuesday. People who bought on Wednesday, either were empty, or didn’t care about price.

Pardon the off topic aside: Upon reading the above, I was surprised to learn there are gasoline pipelines all over the country. My region is loaded with pipelines carrying natural gas and crude, but not gasoline. Never heard of this before… Learn something new every day.:slight_smile:

Gasoline pipelines are quite common in the United States. Refined products pipelines go most everywhere

Thank you for the link. I’m quite amazed at the breadth of “refined” pipelines…
From it, I have gathered that around here (SE Ohio) there isn’t a refined pipeline, which explains why we get our gasoline/diesel via river barge, then trucked from a terminal.

BTW: There has been a frenzy of pipeline construction in our region over the past five years or so, and I note they aren’t included on the map from your link, so it must be a little dated. (These are for “export” of crude/natural gas from the Marcellus shale deposit)