Are there any places to do this?

And more importantly - is this viable? Could margin trading be offered in the world of cryptocurrencies? I don't mean established pairs like BTC/LTC but small altcoins.

Naturally this could be difficult due to the immaturity of altcoins. Theres little liquidity, little volume and I assume other unfavourable factors. Although, suppose its your own exchange (and you're not a meta exchange like Bitfinex once were) and you can post a stop loss at exactly the right level and guarantee it executes - as long as you can satisfy every margin call then everything is dandy - surely?

Originally BitFinex would act as a meta-exchange and place orders at bitstamp and other exchanges. This explains the necessity of them having creditors on the site from which to borrow money.
Today they trade on their own exchange only meaning that they don't need to borrow money to open positions. I suppose it makes sense to actually borrow money from someone interested in lending it - but is it necessary?

It is my understanding that the practice of an exchange opening positions without the asset backing is a naked call. How common are naked calls in the real world? Would it be a good idea for a cryptocurrency exchange to offer uncovered positions? What are the possible implications of doing it this way?

Can anyone offer any thoughts and ideas?

2FA is useless if your computer is compromised, that is if all the 2FA does is provide an ephemeral auth token to 'authorise' an action. Since there's no way to know what it is that you are in fact authenticating - it could be the withdrawal that you're seeing on the screen, or it could be that the attacker is manipulating web page content and doing something else with the token you provide.

For 2FA to be secure (assuming the 2FA device is secure), it needs to sign some data that'll only authorize the very specific withdrawal that you wish to make, so:
- destination address
for convenience this could be just the first 10 characters for example, just enough so that it is inpractical to brute force it in a reasonable time frame using vanitygen
- number of coins
in case the attacker is able to both infect your machine AND socially engineer a scenario in which you willingly send money to an address he controls, except he'll adjust the number of coins once the victim authorises the tx and destination address.

I was looking at http://dustcoin.com/ and noticed that mining some scrypt coins can be twice as profitable as mining litecoin.

Suppose that a litecoin pool operator actually gets his users to mine dogecoins or bbqcoins. He then uses the proceeds to purchase litecoins and pays them out to his miners. The end effect is that the pool operator is way better off than he would be otherwise.

Is it possible for this to be happening seamlessly without the miners not realising? When a scrypt pool miner is issued a work unit would it be possible to have them mine on a different coin than the user assumes?

If the trojan was written to target the bitcoin-qt wallet then it'll be listening out for the decryption event by some means. As soon as it has your password it'll clear out your wallet - and it'll do it faster than you think you could.

I don't think the SR operators or vendors put their proceeds through mtgox or bitstamp. When they're forced to get ahold of fiat they'll use something a little more suitable to the task - cash, perhaps.

Thanks that answers my latter question.

What about the first question? Was not able to find anything on the matter.

Given blockchain difficulty and primerate how does one calculate the expected interval between mined blocks?

How is block reward calculated?

How do you know they're not making short sales?

There's something like 30k coins worth of bids between $100 and $105. It's tempting to say that this may be the bottom for the time being.

You're wrong. If it does go to $1000 then entry point of $100 or $120 is a huge difference.

$120 gives a return of 733%, $100 gives a return of 900%. That's a lot.

Bought, bit by bit across a period of 60 seconds or so, which is puzzling.

The way the wall would be constantly brought down to kiss the best bid screams 'big scary wall of doom' manipulation attempt. At the same time as buys started coming in the seller had plenty of time to cancel the offer and keep his coins which suggests this may have been a genuine attempt to sell his coins. The other theory is the wall was self-bought bit by bit which is also silly because if this really were the bottom there'd be buyers one way or another, what's 7k coins worth at $126 when we were at $134 few hours ago.

But hey, let's hope the coins were an early adopters who wished to cash out and they're now distributed across multiple buyers.

If only they instead used their expertise to bring about a reliable bitcoin exchange.

There's more USD/BTC available to borrow than I've ever seen. No one is taking positions. Everyone is sat on the fence waiting for something to happen.

Went in long with good bit of margin for the first time this morning. Great ride but also great service. Cheers!

Your transparency is very respectable and you may come to learn through this but selling at the bottom of a rally week was not the right call.

Yeah, that was weird to watch. When the wall came down to $133 I was hoping someone would just snatch it up, teach the guy a lesson until he dumped it and I realized it's just a confused bear with too many coins.

Nice to see bitstamp support!

I'm looking forward to a time when the top 10 bitcoin exchanges handle no more than 10% of the trade each and meta-exchanges like bitfinex allow you to take advantage of the liquidity on all of them.

I'm really curious how you manage to have fiat on the right exchanges at the right time though, especially now that you are handling two third party orderbooks. Would you care to shine some light on that?

Fiat come back!