Strategy position management DSL

[lopatin]

There's a lot of room for improving how strategies are written. They currently have the low level primitive operations that we can use to build up a higher level syntax/dsl for position management. Here's a draft of what such a think can look like. Def requires more thought.

val btcPosition = position("my_pos", "btc")
val btcPositionBitmex = position("my_pos", "btc", "bitmex")

val ethPosition = btcPosition / 2

if (something)
  btcPosition("usd") = ethPosition("usd")

maximize(ethPosition.as("ltc") - btcPosition.as("ltc"))