Using Oanda FXBoxOptions Profitably

I've searched the forum for threads about Oanda's FXBoxOptions, but have only found limited information. I did find one of capitalist88's great posts about the calculation of Oanda's effective risk spread, which remains obviously high. Given that these options are clearly not fairly priced, is there a way to trade profitably with them, or are they simply another way for an FX broker to profit from retail traders' penchants for gambling?

Theoretical conjectures are welcome, but if anyone actually has experience using them in a profitable way, I'd especially love to hear about it.

Here are notable previous threads concerning Oanda's FXBoxOptions:

1. Successful Box Option Trades (Oanda)
2. Oanda Box Options
3. Box Option Trading
4. Anyone trading "Box Options" with Oanda???

This is a co-incidence.

About 2 days ago I logged into my Oanda Demo for the first time in probably 12 months. Was playing with box options.

I DO have idea on how you could play them .

The only advantage I can see getting from them is when price stays totally flat. Any other use wouldn't be any advantage.
You pay a maximum price for the chance of missing/hitting the box. This is no different than setting a stop loss etc. In the case of setting a LONG time frame box it appears there would be an advantage as far as risk to reward goes. IF you expected price to not trend soon. Though usually in this case you know there is a range there to trade. A tight range that is useless due to spread costs may benefit maybe.

Like you. I'm willing to bet it's all about the houses chances of improving income. LOL

I intend to look further into it, of course. Maybe a pivot based strategy?

Take note of what Phil McGrew and DarkStar are talking about in that first "Successful Box Option Trades" link.

There is no doubt that any kind of reasonable movement in price would make these a "rip off". Typical of brokers when introducing new types of trading instruments.
Surely the only possible advantage IS during a flat market. But, again, this would need to be carefully looked at.

unless "box options" have changed in the past year, they are an even bigger trick than this "hedge" feature we have been discussing. again it comes down to market mechanics, and the widespread lack of understanding for them. im an options expert, but you dont need to be one to realize why these box options are bad...

last i heard, you could not buy and sell box options from anyone except oanda. this is a FAKE market. at the blackjack table, you can only bet against the house, not other players at the table. this is GAMBLING because the house sets the odds, and they are the only one you can bet against. same with oanda box options...you can only buy and sell from the dealer (aka oanda aka the house). i cant buy a box option and sell it on the open market, because there is no open market, i can only sell to oanada who has a calculator ensuring that the odds are in their favor. this is completely bogus and not even trading, its gambling. if oanda has changed their policy to allow selling of the options on an open market please do let me know.

i really hope we dont get a bunch of people justifying the viability of box options by discussing the strategy they use (like keeps happening with the hedge feature). the strategy doesnt matter, the box options are inherently disadvantageous and therefore any strategy that uses them can be improved by substituting the use of the options with other positions.

I'm not sure that's exactly the case. The box options seem like they actually do provide a decent, though skewed, form of hedging. Certainly, you are right - it would be preferable to purchase real options from ISE or CME because of real-market fair spreads and resale possibility. On the other hand, it's very convenient to have the "options" available all in one platform.

Let me give an example of a strategy I'm testing, and hopefully you can show me what's wrong with my idea. In the chart attached to this post, you can see what I believe to be a USD/CHF uptrend. It seems to have recently broken some resistance, so I am going long. I have a roughly 150 point take profit and stop loss. I usually shoot for a 2:1 TP to SL ratio, but that's another thread altogether. Anyway, I'm aware that the USD/CHF is in a long-term down trend, and that there is intermediate resistance and a possibly floundering US economy on the way, so I want to hedge the downside risk of this trade. I estimate that if the trade were to go south, it would do so sometime in the next week or so. To hedge this risk, I've bought a box option for $55 that pays $94.88 when hit. The stop loss of my trade is set about 144 pips from the entry, with a position of 6400 USD/CHF, for a possible loss of $92.16.

The idea here is that if the trade did go against me, I believe it would go relatively quickly. If it goes down near my stop loss in the next week, I will automatically receive $94.88, offsetting my possible loss of $92.16. If the trade doesn't go against me, I stand to make 150 pips * $0.64 = $96. Subtracting the option price ($55) from $96, I have a profit of roughly $41 (plus a small amount of interest).

Of course, this strategy assumes that the USD/CHF will go one way or the other. It could slide into a trading range, which would throw off the whole scenario. In that event, I could buy an option directly in front of the price action, but that still might not make up my total loss on the original option purchase.

So that's my trade idea. Please critique, and let me know the potential pitfalls of this kind of hedging.

i cant really comment on your strategy, and i have no judgement on whether it is profitable or not. but at the end of the day you are gambling with this feature. you are betting against the player who fixes the odds of the game, only luck can save you

Hi Drummer,

I'm the one that did that "back of the envelope" calcs on the odds of these box options. I also did create a strategy for using them, but dropped it because I didn't like the high drawdown potential. I actually made a little money with the strategy, but I'll make an educated guess that I would have lost in the long run because my edge was so slight and the drawdowns would have killed me. I think my motivation was more to be kind of "in your face" to Oanda with respect to these so called "options" by "out-statisticing" them (yes I invented a new word there ).


Anyway, the strategy involved using multiple regression analysis on four variables which described the price action of two completed weekly bars. The output of the analysis gave me a projected close for the next week which I used as a starting point in the next step.

The next step was to use a spreadsheet that I had created along with some historical data to determine the historical probability that the week's close would fall between two points that were equidistant from the projected close. So if I had a projected close of 1.4000 on EUR/USD for example, I would check the probality that price would fall in the 100 pip range from 1.3950 to 1.4050.

I had to find a a "hit range" like this at the end of the week for the box, set up in such a way that the trade would have a positive expectancy. The idea was to set up the box for the last hour of the week (or whatever the thinnest possibility was; I forget) which stretched from the high to the low of the "hit range" (more on how I found this in a second). So at the beginning of each week, after the spreads narrowed and the box prices dropped, I would set up these high thin hit boxes at the end of the day for the coming Friday.

I had to find a "hit range" so that expectancy was positive based on the probability of a hit and Oanda's payoff. So for the example above, if the option would cost me $50 and paid $55 then I was risking $50 to gain $5. To get a positive expectancy my probability of a hit would have to be 1-5/55 or about 91%. If the probability showed up at 80% or something then I would have to stretch out the hit range and retest based on the new Oanda payoff. I had Excel set up so that it calculated probabilities for a whole array of possibilities so I could check them quickly against the payoff values in Oanda's platform.

The key to the thing was that I wasn't centering my target zones on the last close, but was instead skewing them up or down based on the multiple regression analysis that I did at the beginning to find the projected close. As I said though, I think this strategy would have been killed by drawdowns, OR Oanda would have simply made the payoffs even worse than they were to compensate for geeks like me. (Being very undercapitalized, I'm still hunting for a part time job to pay the bills but I keep putting it off; anyone know of a successful trader who might want to hire a cheap geek?? )

I'm doing this off the top of my head so if I screw this up don't TAZE me bro!!

There are multiple possibilities here depending on whether the box gets hit or not, the stop gets hit or not, the TP gets hit or not, each with its own payoff and probability, so the expectancy of this thing would be hard to figure out to say the least.

1. Nothing gets hit: the payoff is (exit - entry - $55)

2. Box hit only: (exit - entry + $40) I'm rounding the payoff of the box to $95, and since you paid $55 the net is $40.

3. Stop hit only: (-$92 - $55) = -$147

4. TP hit only: ($96 - $55) = $41

5. Box and TP hit: ($96 + $95 - $55) = $136

6. Box and stop hit: ($95 - $92 - $55) = -$52

7. TP and stop hit: not possible

8. All three hit: not possible

So there are six possibilities, each with its own payoff and probability, and I have no clue what the probabilities are obviously. But one thing to keep in mind is that EVERY possibility includes a $55 cost because you've already paid for the box.

Ok, now in order to figure out if box options make sense, we can compare your strategy to another one which doesn't use box options. I'll have to make a bunch of assumptions, but if we just make the same assumptions in both models, with the only difference being the presence of box options or not, then we should be able to isolate the "box option effect" so to speak.

So just for chuckles, let's just pretend that all the probabilities are equal so we can dispense with that question. I know they're not equal, but it doesn't matter because we're using the same assumption in both models. I'll also assume that in the cases where the payoff depends on your exit (neither stop nor TP gets hit) that the exit is at the entry price. Again, same assumption for both models just to "eliminate" that variable from the analysis.

So...we add up all the payoffs above:
(-55 + 40 -147 + 41 + 136 - 52) = -37

But don't despair because this number is based on unrealistic assumptions so it's not saying that your strategy has a negative expectancy. It's just an index for comparison. Now let's go through the same process without box options.

The possibilities and payoffs are:

1. Nothing gets hit: payoff is 0, (assuming exit = entry like we did in the last model)

2. Stop only hit: -92

3. TP only hit: 96

4. Both hit: not possible

These add up to 4.

So in conclusion, everything else being constant, the addition of box options to the strategy has a negative effect of (-37 - 4) = -$41.

Conclusion: Merlin is right; you can create a strategy that implements your market opinion without box options and it will save you money in every case.

Q.E.D.! (LOL, I love saying "QED")

Merlin would also have in mind real options trading.

When you add the advantage of being able to write your own calls and puts too, you create a whole new range of trading strategies.

You're right, he mentioned that he would be interested if they ever change the feature into an actual product that could be exchanged with other traders in a real market.

I wouldn't though. I use Oanda now and used to trade options years ago, but I already have a tendency to make my methods too complicated as it is without them.

By the way, I played with these last night. For about 5 - 10 minutes. Realised I'd done it before (long time ago) and came to the same conclusion. "NEXT!!"

capitalist, i am really starting to dig you man. instead of being lazy like me, you went ahead and did the actual math proof. i lazily skipped that part, trying to make my point only with the logic. thank you sir, you are a shinning light on this forum. and im not just saying that because you said i was right LOL (i already knew that)

yes, out-statisticing them is the only way to win. but you'd have to believe oanda is pretty darn dumb for that to work. options valuation has come a long way in recent decades, and the math is pretty much solid these days (ie black scholes etc). when i first heard about these box options i did some math to see if they were mispricing the options at all, just for kicks. unfortunately oanda is no dummy.

sombody throw drummer another vouch so we can get him back in this discussion!

I just added a vouch. He also started a thread on neural nets that I think is going to ROCK!


Thank you for your comments too Merlin! I'm actually still just an undercapitalized doofus in an apartment somewhere.

Thanks, Merlin (and capitalist). I appreciate it. Capitalist, you were spot on about the price analysis. Silly me, I totally forgot about that $55 purchase price! Should have known it was too good to be true. It looks like to create a profitable hedge using Oanda's options, you'd have to know just where the trade would go if it went against you, as a large box option as a hedge does not provide enough return on investment to cover most trades. Basically, to win with these options, it seems like you have to have a valuation model that's way more accurate than Oanda's... we're talking 15% better or more, as Oanda's egde (ie, risk spread) is huge.

Too bad. Maybe I'll play around a little more with them in small amounts and see if I can find some clever strategy. I mean, hey - they could use the money anyway... 'cause once my NN is up and running, their 0.9pip EUR/USD spread won't have nothing on my statistical edge

Have a good weekend guys.

famous last words

glad we could enlighten you man remember...shit in, shit out! LOL

FXBoxOptions my frigid Chicagoan ass

oh my we are having way too much fun lately

Wait, that's a different thread... are we lost??