1400 days ago, I started building an
algorithmic trading platform. Today, I
executed my first real trade with it.
Update: Dec 4th, 2024:
For a step-by-step guide on deploying your
algorithmic strategy to the cloud,
check out this article.
Four years ago, during my Masters, I drew my
first diagram for “the Money Printer”.
I had no idea what the future would hold.
At the time, “algorithmic trading” was barely
more than a buzzword to me, like jargon such as
“no-code”, “blockchain”, or “artificial
intelligence”.
The first diagram I made for “The AlgoTrading
System”
But my education at Carnegie Mellon taught me a
lot and I began to see the true value of
algorithmic trading. I
learned how “AI algorithms” actually worked
under the hood.
More importantly, I saw how these models were
applied in the real-world. When I had my first
big-boy internship, my first project was
integrating Google AutoML for automating
document classification and triaging. This
simple project eliminated the need for manual
document tagging, saving the company over
$400,000 (as an intern project)!
This experience made me realize how much value
AI can bring to businesses. Eventually, I
started integrating AI features into my trading
platform, such as using genetic algorithms for
trading strategy optimization, or using large
language models to create strategies using
natural language.
And now, 4 years in the making, I’ve executed my
first real trade on the world’s most powerful,
no-code algorithmic trading platform.
Here’s why this is so important.
My first real trade with my Alpaca account
To start, what is NexusTrade?
Before I dive into why this is such a big deal,
I need to give you some context about the
NexusTrade platform.
NexusTrade is the world’s most powerful (and
easiest to use) fully automated trading
platform. It allows anybody, even you or any
non-technical beginner, to deploy nearly any
trading idea they can possibly imagine, using an
interface that we, as humans, normally
communicate with — language.
Creating an algorithmic trading strategy using
natural language
By using natural language, NexusTrade users are
able to interact with the market in a
fundamentally unique way. This doesn’t just
include creating and testing trading strategies.
For example, this also includes giving users the
ability to perform detailed financial research
to inform their investing decisions.
The average return for SPY after thanksgiving
from 2018 to 2023
By “financial research”, I mean things like:
-
Fundamental analysis: using a stock’s earnings statements to
measure its financial health
-
Stock screening: finding novel investment opportunities
based on technical or fundamental data
-
Trend analysis:
using objective technical and fundamental data
to unveil hidden patterns in the stock market
This type of automation and analysis was
absolutely inaccessible to the average retail
investor, but it is exactly what the quants at
places like Jane Street and Two Sigma do
everyday. Before NexusTrade, retail investors,
quite literally, had no way to perform this type
of financial research.
More importantly, they had no way to act on
these insights. While the analysis features are
powerful (especially for a savvy-investor with
years of experience), the trading strategy
automation aspects are what makes NexusTrade so
unique.
What
EXACTLY do I mean by “algorithmic
trading”?
As I mentioned in the beginning, words like
algorithmic trading can seem like jargon.
However, it’s surprisingly very informative.
Algorithmic trading refers
to the process of using algorithms to execute
trades. An algorithm is like a baking recipe –
it is a series of steps that you take in order
to achieve a desired outcome (like baking a cake
or executing a trade).
Put simply, NexusTrade is a platform that allows
anybody to create their own “series of steps”
(also called a
trading strategy) to execute
rules in the market.
Here are some of the types of rules you can
create with NexusTrade.
Simple rules executed at regular intervals
Most brokerage platforms allow users to create
simple trading rules such as
Dollar Cost Averaging. An
example of this is a strategy like the
following:
Spend 5 percent of your buying power in SPY
every 14 days
A simple trading rule that purchases a
market-wide ETF every 2 weeks
Strategies like this are useful for investors,
but they’re not exactly innovative. Within
NexusTrade, they’re valuable to demonstrate an
idea: a strategy is simply a set of rules for
executing trades.
In this example, we will spend 5% of the
portfolio’s buying power buying SPY every 2
weeks. We can see that the green line (our
strategy) significantly underperforms the grey
line (buying and holding SPY). This is both in
terms of the raw percent gains this year, and
the risk-adjusted returns (as shown by the
sharpe and sortino ratio).
Because of these results, we can conclude that
we don’t want to deploy this type of strategy
unless we’re particularly risk-averse and
predicting some form of market downturn.
This information is informative and not easily
demonstrable with traditional investing
platforms. However, what makes NexusTrade so
powerful isn’t the ability to create simple
strategies like these. It’s the ability to
create sophisticated ones.
Multiple trading rules using complex logic
An example of a complex set of trading
strategies
In addition to simple trading rules, NexusTrade
allows you to create portfolios with
sophisticated trading strategies. The screenshot
above shows a strategy set called “Buy TQQQ and double down when positions are
down. Sell a small portion when positions are
up.” Let’s take a look at the rules:
A screenshot of the rules on the UI
-
Buy 50 percent of buying power in TQQQ Stock
under specified conditions
-
Buy 25 percent of buying power in TQQQ Stock
under specified conditions
-
Buy 10 percent of buying power in TQQQ Stock
under specified conditions
-
Sell 5 percent of portfolio in TQQQ Stock when
(# of Days Since the Last Filled Sell Order
> 30) and (TQQQ Positions Percent Change ≥
10)
To understand what “under specified conditions”
means, we have to zoom in on the condition. For
example, by clicking the “Buy more aggressively when starting” button, we can see the 3 conditions. They are
-
# of Days Since the Last Filled Buy Order >
7 AND
-
Portfolio Value < 17000
AND
-
(Position Value < 100) or (TQQQ Positions
Percent Change < -20)
What this means is, we will trigger the strategy
to buy 50 percent of buying power in TQQQ if
every one of these conditions are
triggered.
Similar rules exist for the other 3 strategies.
They dictate exact rules for when the strategy
should execute actions in the market.
The beauty of this is that it shows
exactly when we will enter
(or exit) a trade and why this will happen. It’s
also replicable, which changes trading from an
art (as demonstrated on social media) into a
science.
Finally, the UI also shows the historical
performance of these rules in different market
conditions.
For example, while the above screenshot shows
the performance of this strategy in the past two
years (during a bull market), we can change the
backtest to show us the performance of this
strategy from 2020 to 2022.
Backtest results from 2020 to 2022, the
portfolio is down 40% even though buying
and holding SPY was up 12%
The statistics next to the backtest affirms
this. It shows that while the percent return is
much higher (or lower) than SPY, other
measurements, such as risk-adjusted return and
drawdown, are much worse.
Being able to extract these insights and test
out different algorithmic trading strategies is
extremely valuable all by itself.
However, today, I made a huge leap forward, by
executing a real trade directly in the
NexusTrade platform.
Executing a REAL trade with NexusTrade
My Alpaca Portfolio connected within
NexusTrade
While the ability to create and test trading
strategies is useful, up until now, its been
solely a paper-trading platform. What this means
is that while you could deploy your strategies
to the market, you were trading
monopoly money to get
practice and learn how algorithmic trading
works.
However, I’m excited to share that at
approximately 7:25PM EST, I made my first real
trade with actual money
using NexusTrade.
Creating an order with the NexusTrade app
While this trade was
manually triggered
(so not executed by one of my trading
strategies), finishing the integration with
Alpaca is now almost entirely done.
The architecture is built, and the plumbing is
implemented, and fully
autonomous
trading is within grasp for every single retail
investor in the world.
And you can get started right now by
clicking sign up or clicking the link
below.
Why this matters?
This launch is a major milestone for the field
of finance. It is the first time in human
history where anybody can become a quant, and
ultimately learn how to make money in the stock
market for themselves and for their family.
NexusTrade isn’t a black-box buzzword machine
that promises ridiculous returns with its “AI
Algorithms”. It is a tool that anybody can use
and learn how to trade with.
It facilitates financial research, helps people
bring their ideas to life, and does so in a UI
so easy that my 13 year old niece can
understand. This makes it easier than ever
before for people who want to become profitable
in the stock market.
And this is only the beginning. With the
integration of live-trading, more possibilities
are unlocked. For example, people are going be
able to share their trading ideas within their
social circles in an objective, systematic, and
replicable way.
This increases financial literacy and
accessibility because you can clearly
articulate how you are making money in the
stock market.
Moreover, this sets the bar for future investing
platforms.
It’s not enough to claim to use AI; you have to use AI to bring real value to the
people using your software.
My journey through Carnegie Mellon and my past
internships has helped shape NexusTrade into the
platform it is today. Early in my career, I’ve
demonstrated the value of AI, even as an intern.
Now, I’m showing how it can be used to make
algorithmic trading accessible to billions of
people.
I cannot understate how big of a deal this
actually is.
