Second guessing: This could very well be an algo who is sucking these 10c off each trade...
See an example for that "trading model" (buying at the ask and selling at the bid) here:
http://www.nanex.net/aqck2/3525.html
True - but I don't think this is it either. I called Kitco and asked where they get the bid ask from. They said Reuters. This is inconsistent with what I was looking at on Bloomberg. My Bloomberg terminal does not post bid and offer prices, just the last trade.
I'm a commodity trader so I look at this type of stuff all day long. The "bid" price they were quoting was simply the last trade on the exchange for the front silver contract. The "ask" price was simply this plus 10 cents.
My experience is that bid/ask spreads in a market with many participants are rarely always uniformly spaced in time. For instance, you have many buyers and sellers competing for a price. Dealers are constantly posting their 2-ways which usually are uniform. At some point or another an axed buyer or seller will "cut" a dealer's market, robot or not.
I find it hard to believe that after trading for size on a bid, that the bid stays put. Usually, when dealing, if you get done at your bid, and your two-way is 10 cents wide, you would make it 10 cents wide around the last trade price. I.e, you bid 29.97 for 100 lots and you get them, your price is now 29.92/30.02. This was not happening on kitco's site.
It also makes sense, given the current metals environment, that they have many more buyers than sellers. Furthermore, another PM website (Liberty coin and precious metals) was reporting a different, tighter and more reasonable bid/ask at the same time (priced around the last trade).
I guess, at the very least, this will bring attention to the issue. Frankly, I DO NOT TRUST THE INFO POSTED ON KITCO.
EDIT: The other alternative is that Reuters has crap information. This would not be out of the question either. It does not change my stance on not trusting the bid/ask spreads on Kitco though...