# Understanding Arbitrage Strategy (problems and solutions)

I think it's easier to understand with an example:
Suppose this arbitrage between 3 markets:
BTC/USD = \$1000 USD // just to make easier the number
BTC/EUR = \$870 EUR
EUR/USD = 1.17 USD

We buy 1 BTC at \$1000, then sell it for €870 and then change them to USD by \$1017.9 USD.

This looks like a non-risk always win strategy, which is obviously impossible but I don't fully understand where is the problem.

I think possible problems are, speed, liquity and commissions.

a) Speed: prices will change with time so we need to do the 3 transactions at the same time, is this really so difficult with current computers?

b) Liquidity: to execute orders quickly we need market orders, so a big order will move the price. But then we could avoid this just moving a small qty (like \$1000 on the example)?

c) Commissions: we could just trade when the profit expected is greater than this. In our example if commission is 0.3% per transaction it will cost \$9, so we end with \$9 in profit instead of \$18, but still being a good number (0.9% profit)

What is the problem with this arbitrage strategy? because is naive to think it will work of course but I can't see where is failing.

• What evidence do you have that such a price arrangement has ever existed? If I can say "abracadabra" and a bar of gold appears out of nowhere, that is also an always-win strategy. The problem with it is that it doesn't occur in reality. Where did you get your numbers? Do they occur in reality? Jan 6, 2019 at 18:32
• A single bitcoin transaction can take hours (especially if you don't want to pay transaction costs). The bitcoin price is likely to change in the mean time. To make all 3 transactions at the same time you need to already have a bitcoin that you can sell on a euro exchange while buying one on a USD exchange and trading EUR for USD. Jan 7, 2019 at 18:17

If that situation ever existed where I could turn \$1000 into \$1009 after commissions guaranteed with a total round trip time of seconds I would do that all day long, and so would everybody else.

The people who can trade millions of dollars at a time would do so. They have programs looking for these opportunities. That window of opportunity would be shut so fast almost nobody could take advantage of it. It certainly wouldn't stay open long enough to alert your friends.

The problem with this strategy is that you made up the numbers and assumed nobody would notice that opportunity.

• Yes I know is naive to think I can make money like that, was just an example to understand why we cannot. Because I'm pretty sure there are bots out there earning money with arbitrage. Is the time then the problem? I mean the one using the fastest bot is the only one earning the arbitrage? Jan 5, 2019 at 19:27
• You can see some examples here (search bitcoin arbitrage): tradingview.com/wiki/Spread_Charts Jan 6, 2019 at 22:47

I am not familiar with the BTC market so I cannot offer a specific answer for that security.

In the option markets, arbitrage opportunities frequently arise (conversions, reversals, box spreads, etc.). I have done some but there are several obstacles in retail's way.

The big boys have programs looking for these situations. The window of opportunity is small.

Retail's commission structure tends to exceed the guaranteed profit.

The size available is often small, making flat fee commissions take a larger bite out of the profit.

In today's era of HFT and quote spoofing, what you see isn't always what you get. There are times when I hit the bid or the ask and as fast as I click, price moves. If I move my price away from the market or cancel the order, the attractive price returns only to play hide and seek every time I re-place the order (illiquid securities).

Back to BTC. If I had to guess at a legitimate reason why your arb strategy isn't actually what it appears to be, I'd say that you might not be evaluating the components at the respective bid and ask. If the prices were:

\$1,000 x 1,011 USD (last trade \$1,000)

\$864 x 870 EUR (last trade \$870)

• then the arb would exist at last trade prices but would not at the actual bid and ask prices.
• OK, I understand so you say sometimes arbitrage opportunities could be also a trap from faster/smarter bots. They create a fake bid/ask price (spoofing) which is impossible to catch with our computer speed. Jan 5, 2019 at 19:23
• When I've chased arb positions (options), before opening the first leg, I have also set up a closing trade for that first leg in case the second leg isn't available due to market movement. Have all orders set up before executing order number one. You don't want to be the deer in the headlights munbling "What happened" ?? . And FWIW, it's really not worth the effort. There are easier ways to make a buck. Jan 5, 2019 at 19:58