Is profitable Lindex trading still mathematically possible?

I hesitated for some time before posting here (about a month) because I figured I had to have gotten the math wrong because I hear all the time about people being able to make small profits off the spread on each transaction but is it just me or is this not possible anymore and if I'm wrong please correct me because this is driving me nuts. I'm not so much interested because I'm thinking of doing this, it's more of me trying to figure out an academic problem that I can't seem to solve and there is almost no documentation on currency trading.

I didn't put any actual sales through but here's the Lindex estimates which from previous experience seems to be accurate to about US$0.01 or so.

Sell L$1000 - Market is 269/1 and the floor (with a couple million up for sale) is 259/1 which means that I'd have to sell at 260/1 for an estimated Proceed of US$3.85 before fees and US$3.72 after fees.

If I were to turn that around the market buy is 260/1 and the best rate to buy is at 268/1. The estimated cost for that transaction is US$4.04 which would be a loss of approx. 19 cents (US).

You could push my limit sell figures one cent in either direction but you'd have to fight against anywhere from 1.5 million L$ to 6 million L$ which make a fast turn around anywhere from unlikely to impossible, it also would not make much of a difference since you'd still have a large loss.\

I know that there is also a technique of trading on shifts in the margin but since the shifts are usually only one cent because of the Linden blocks I don't see how that could be profitiable either because of the issue I brought up before, that and the people who supposedly do that successfully don't like to share their secrets.

Any insights on this would be greatly appreciated. Academic curiosity has gotten the better of me and I have to assume I'm making a mistake somewhere.

If you sell 100k at 259 you get roughly 372.11 after fees.
If you then turn around and buy at 269 costs you 372.05 including fees.

That is the roughly break even point on a 10 point spread. Using the 10 point spread you can do roughly 5 complete buy/sell cycles in a month.

So yeah it is POSSIBLE to come out ahead trading. TO make any kind of real profit though you would have to be doing flips (full cycles) on an absolutely obscene amount of money.

When you sell on the LindeX, you pay a fixed percentage, so the cost as a percentage of the transaction amount never changes. If you sell $100, you pay $3.50. If you sell $1000, you pay $35.00. (For this purpose, it doesn't matter whether $ means USD or L$.)

When you buy, you pay a fixed dollar amount, $0.30, USD. This means the cost does vary as a percentage of the transaction amount. If you spend $1US on a purchase, a whopping 30% goes to the fee, leaving you with 70 cents worth of L$. But if you spend $100US, only 0.3% goes to the fee, leaving you with $99.70US worth of L$. Change it to $10,000US and the percentage drops to 0.003%.

Now let's plug in the exchange rate spread, but let's use easy round numbers instead of the actual exchange rate. Suppose the exchange rate for buying is that $1US gets you L$1000, before fees are added, so with fees it would be $1.30US gets you L$1000. If you wanted 10K L$, it would be $10.30.

For the spread, let's suppose it's a 4% spread. In other words, if $1US buys L$1000, then you only need to sell L$960 to get $1US - before fees. So if you sell L$10,000, 350L$ goes to the fee, so you're really only selling L$9,650, which gets you $10.05 USD. You've lost 25 cents US.

Now increase the quantity by a factor of 10. The selling fee goes up, but the buying fee doesn't. It costs you $100.30 US to get L$100,000. When you sell, you lose L$3,500 to the fee, so you're selling L$96,500, which gets you $100.52 (96500 divided by 960). You've made 22 cents US. So clearly there's a breakeven point for size of trade, below which you lose money and above which you make money.

That's with a 4% spread. When the real best buy rate was L$270/1USD and the best sell rate was L$259/1USD, the spread was 11L$, or as a percentage, 11/270, which is about 4%. The buy rate has slipped but the sell rate hasn't, so we now have a spread of 10L$, or as a percentage, 3.7%.

Sticking with the percentages, but using my round numbers, you'd have to sell L$963 to get $1US before fees. Remember that before, to make a profit, you had to spend $100.30 to buy your 100K L$, and that hasn't changed. And now, when you sell that 100K L$, you pay the same 3.5% fee, so you're still only selling L$96,500. But the exchange rate is worse, and you only get $100.21 (96500/963). You've lost 9 cents US.

Bump the number by another factor of 10. You spend $1000.30 to get 1M L$ - the 30 cent fee is still just 30 cents. After subtracting the 3.5% sales fee, you're selling L$965,000. Divide that by the new exchange rate of 963, and you get $1002.07. You've made $1.77. It's still profitable, but the breakeven point is now higher than it was before.

Here's another way to look at it: If you start by spending 1USD, your round trip fees are 33.5% - 30% on the buy (30 cents of the dollar) plus 3.5% on the sale. At 10USD, the round trip fees are 6.5% - 3% on the buy, 3.5% on the sell. At 100USD, the round trip fees are 3.8% - 0.3% on the buy, 3.5% on the sell. That's why with a 4% spread, you can make money at $100, the spread is still larger than your fee. But with a 3.7% spread, you lose money at $100, since the fee of 3.8% is higher. At 1000USD, the round trip fee is only 3.53%, so you can still make money with a 3.7% spread.

One last thing - this ignores the fees, if any, for actually converting your USD balance with LL into dollars in your bank account.

The interesting thing is that by discussing this matter you are just making it less profitable because more people will become aware and start trading.

Anyway
11791926 Limit Sell L$1,350,000 L$1,350,000 L$259 / US$1.00 2009-03-29 20:57:26 2009-03-30 20:42:35 fill US$5212.40 US$5029.58
11716721 Limit Buy L$1,350,000 L$1,350,000 L$269 / US$1.00 2009-03-26 06:49:46 2009-03-29 18:56:07 fill US$5018.57 US$5018.87

Yes and no Elanthius. Yes people learn it is possible BUT how many are really willing to risk a thousand US$ or more to make a few pennies? The fact is that any sudden shift in the lindex and you can lose quite a bit of money. So it's not a game for those that are not willing to lose whatever they have to trade.

Elanthius, think back to our previous conversations on the forums, when have I held back to protect the happiness or profitability of others when deciding what I talked about or linked to for that matter. That's not to say that I don't care about what people think, I just don't let it effect me to the point where I self censor to protect the interest of others.

Darkness, using your numbers if your limit was high enough it seems like you could replace volume with quantity of trades but that's a lot less efficent and would require a lot more turnarounds as well as impacting their profit margins since they'd still be paying $0.30 on the turnaround. If you had less capital though someone could go down to making 1 cent per transaction and just keep making transactions as fast as the market will allow. Overall they'd make a lot less unless their limit was 6x more than the other person (assuming 1c profit per transaction instead of 6) but it's an interesting idea.

Been aware of the possibilities and have certainly had the $L flow to attempt it... but there's a whole lot to be said for not messing with one's reserves in any way, shape or form.

I could run the exact same #'s as Elanthius with 11 days of saving up, but... meh.

Such conservatism saved my tail when it came to the openspaces, also. True, there's not a whole lot of risk to currency arbitrage that I can see... is it essentially 3000 USD for minimal effort yearly? Maybe.

But still the best leverage I ever got was out of a good rep and ridiculously restrained conservatism.

Other people's mileage may vary. Anshe Chung did far better than I ever shall; I'm sure Elanthius is doing better than I am too. And yes, that takes guts and vision.

But on the other hand... I know for a fact I'll never lose my shirt because I was doing currency arbitrage and something went wonky.

Quite true. It's not at all clear that tying up $5000 in order to make $10 every 3 or 4 days is worth it. If I'd bought and flipped some land with that money I surely would have made a lot more profit. Also, as you rightly say, there's a lot of risks that the LindeX will do something crazy and even a couple of point move could wipe out weeks of trading.

It does have the added benefit that the cash is still quite liquid. i.e. at any time I can cancel the trade and I have the actual cash in hand. Only thing is you can't be sure if you'll have L$ or US$.

That used to be very true. Long ago the spread was 13 points and you could easily turn 3-5 round trips A DAY. The problem is now the amounts at the main price on the lindex are so huge that you really can only do 4-5 round trips a MONTH. No way to speed it up unless you can work on a 9 point margin which I have never even run the numbers on but suspect the amount of cash tied up would be even more terrifying.

You cannot make money under any circumstances with a 9 point margin. 269/260 = 1.0346, which is less than the 3.5% fee that linden labs charges for a sell.

I'll start making little signs, 10 point margin or bust