//THREAD

Let's talk fees.

One of the common phrases on CT is that "fees kill your gains"

Most humans are visual learners and the ones who aren't usually need some sort of proof to back up a statement for them to believe it. Normally no data is presented.

Time to change that
Just so that I don't lose anyone a taker fee is the fee you pay when you enter or exit a position with a market order. This fee is 0.075%

A maker rebate is the rebate (rebate=money you get back, think a credit card cashback) when you use a limit order. The maker rebate is 0.025%
So if I market in and out of a position I will pay a -0.15% fee.

If I use 1 market order and 1 limit order I'll pay a fee of: 0.075-0.025=-0.05%

if I use limit orders for both entry and exiting I get a cashback/rebate of 0.05%
So time for some charts. Below are some simulated equity curves I created.

black = no fees or rebate

red = 2x taker fee

orange = 1x taker fee 1x maker rebate

green = 2x maker rebate
Obviously the more fees you pay the worse your equity curve.

In this example you can see if you would enter every trade with a limit order and exit with a market order you'd be at around breakeven

If you would've marketed in/out you would have lost 1/3 of your capital
So now that we've seen some charts let's see some data on how much our returns differ between 1x maker and 1x taker as well as 2x maker and finally 2x taker vs no fees.
First with 100 trades:

2x taker: 13% loss vs no fee

1x taker: 7.2% loss vs no fee

2x maker: 4.85% gain vs no fee
Secondly 250 trades:

2x taker: -31.2% loss vs no fee

1x taker: -17% loss vs no fee

2x maker: 11.7% gain vs no fee
Finally 500 trades:

2x taker: -52.8% loss vs no fee

1x taker: -31.3% loss vs no fee

2x maker: 22.1% gain vs no fee
So now hopefully you can see the true impact that fees can have on your gains.

So now that you know fees actually *do* have a significant impact you maybe won't lose out on as much of your profit due to overlooking a seemingly small part of your trading.
Method:

So what I did was I created a short script that creates randomized equity curves with returns following a normal distribution. Not perfect by any means, but good enough to illustrate our point.
For the histogram of return differences all I did was take the percent difference between the non-fee final return and the different equity curves.
Here is a real life example from @7ommyZero

If you do not account for fees or use some sort of software that can show the difference between returns with and without fees you should
ADDENDUM:

Does it really help to use a ref link? Let's investigate. Affiliate links help you save 10% on fees*

*On BitMEX
Example, 100 trades:

(black line = fees paid with affiliate discount)

With 1x taker fee you would have paid 9.8% less in fees

With a 2x taker fee you would have paid 9.4% less in fees
is a 9.5% difference for fees really that big though? Yes, yes it is. In the long run the money you save from fees racks up.

If you were to execute a 100USD trade everyday for 1 year with 2x taker fee you would pay 55USD simply in fees!
If you would've used an affiliate link you would've paid about 50USD all in all.

This may not seem like a huge difference, but once you're trading big amounts that might be 100K paid in fees vs 90K paid in fees.

Which means you would've saved $10000 just by clicking a link.
So that's why I'd recommend you use an affiliate link for all of your exchange accounts.

If you'd like to support me and save fees when you're trading I'd be very happy if you decided to use my affiliate link for BitMEX:

https://www.bitmex.com/register/nNsnrn 
Yet another real life example. This time of a backtest made by @spicyofc

Brutal.
You can follow @CosmonautC.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: