We have reached a new record high this week due to some more volatility.
Notice that this approach works best when there are some significant corrections and volatility, due to increased scalping activity.
Note also that we are using
only 38% of the available funds at the moment, just to be able to ride a better possible large move and let the "platform load up" on these occasions.
The current PNL is around
670K after
574 days, starting with 2MM. This means that if we close right now all position that
is exactly what we have generated in net terms (
net of anything and
excluding interests).
The PNL figure can be decomposed, for accounting purposes, into 2 parts. One is called
Realized and one is called
Unrealized.
The
Realized component is the one resulting from all
current lot matches.
The
Unrealized component is the one stemming from lots which, on the other side (buy/sell), at the moment do
not have a "
matching"
counterpart.
This means that the Unrealized is largely
arbitrary (because the
lot matching can be done, in principle, in an incredible number of ways, talking of something in the vicinities of 2000! x 500! in our case, or in any case
astronomical and unwritable numbers).
Note that the
future profits can only be part of the
Unrealized. Because the
Realized profit component,
is, well, as the name suggests,... "
realized"
Well, to be more precise, not exactly so actually because even the so-called "Realized"
may be formed of pairs that change in time, depending on the algorithm used.
For instance, in case we use some
time-based method, the lot matchings already done are
immutable. But, if we use some criterion based on
price, they
can and will change in time.
What matters is your PNL, the rest is accounting stuff, and
if you want to really
push an interpretation of the
Unrealized, which is arbitrary (in a practically "infinite" way) you can take its
range as an "indication" of future profits. But there are
better indicators for that.
Remember that, as always remarked on any occasion, all illustrations I show on the web are always using
fictitious $$$. On the other hand, the algorithm obviously
cannot distinguish between real or fictitious $$$ and does obviously the same actions.
I am a pure free technology provider. I am
not a company. I am a University researcher and a rather well-known statistician showing the results of his research and possibly engaging in strictly private recreational activities with (necessarily not numerous)
selected friends or investors who like my initiatives and projects.

Here is another example of the result (this is 1MM init cap and
218 days running, always "fictitious" $$$ of course

).
You will in general see a "superlinear" behavior of the returns with respect to used capital.
Whether he closes all now,
or not,
that is the actual net profit ("fictitious" $$$ of course

he sees in his account.
It is almost
never convenient to "close" or stop the
algorithmic dynamic process because the goal, conceptually, is to continuously turn the
Unrealized profits into
Realized ones.
When you close
abruptly, you turn
U immediately into zero and do not allow it to flow into
R.