Koin Capital Inc (“Koin”, the “Firm”, “we”, “us”, “our”), a Delaware corporation, is a
provider of algorithms-based investment advisory services. Koin was formed in October 2020. Jiyu
Ding, Hang Xu, Yang Zhang, and Shuai Hao own 59.48%%, 21.25%, 4.27%, and 15%,
respectively, of all of the outstanding shares of the Firm.
Koin provides algorithms-based, non-discretionary advisory services to retail investors
exclusively through a mobile application developed, owned, and maintained by Koin (the “App
Platform”). The App Platform provides clients with investment recommendations generated by
Koin’s proprietary artificial intelligence algorithms. Our algorithms employ machine learning
methods to analyze various financial factors and market movements and generate investment
recommendations. Currently, all our investment recommendations are limited to the selection of
stocks traded on a U.S. national stock exchange. There are a finite number of individual stocks that
our algorithms may suggest to our clients for their potential investments (See “Provision of Stock
List” below.)
Koin currently provides the following categories of algorithms-based investment advisory
services through its App Platform:
Provision of Stock List. Our algorithms categorize the stocks listed on the U.S.
national stock exchanges according to one of the eleven official sectors identified by the Global
Industry Classification Standard (GICS), an industry taxonomy developed by MSCI and Standard
& Poor’s for use by the global financial community. Our algorithms will generate a list of stocks
that have finished in the top 20% based on risk adjusted returns, or the “Sharpe Ratio”, over the
past trailing one-year period in each of the eleven GICS sectors (the “Stock List”). The Sharpe
Ratio is a calculation that balances a stock’s return with such stock’s risk. Each stock in the stock
market carries different degrees of risk. The goal of the Sharpe Ratio is to take into account the
different degrees of risk each stock has when comparing their returns. The purpose of this ratio is
to help investors determine whether the risk taken was worth the reward received. If two or more
stocks delivered the same returns over a given time period, the one that has the lowest risk will
have a better risk adjusted return. The formula for calculating Sharpe Ratio formula is set forth
below:
Our Stock List which will be refreshed on a calendar semi-annual basis (i.e. on June
30th and December 31st). The Stock List is made available to all of our clients through the App
Platform. If a client so wishes, such client can invest in any stock on the Stock List.
Base Portfolio Recommendation. When a client signs up for Koin’s services, the
client is required to complete a questionnaire in the App Platform that aids our algorithms in
developing a good understanding of such client’s investment objectives. We initially ask each
client to provide his or her annual income, savings, net worth, employments status, investment
experience, and risk tolerance. Collecting such information allows our algorithms to determine a
suitable investment portfolio for such client. In our “Base Portfolio Recommendation”, we have a
second questionnaire in the App Platform that requires each client to pick their risk tolerance and
investment preferences from a list of choices. Each client is asked to identify (i) his or her personal
risk tolerance by selecting one of the three options: “Aggressive,” “Moderate,” and “Conservative,”
and (ii) their preferred portfolio focus area out of several options (such as “Technology Stocks”,
“Cyclical Stocks”, “Defensive Stocks”, etc.). Based on the
information we collect from each
individual through the two questionnaires, our algorithms propose a customized investment
portfolio and determine the appropriate weighting of sector or sectors of stocks in such customized
portfolio (the “Proposed Portfolio”).
The Proposed Portfolio is generated by using five machine learning algorithms (linear
regression, random forest, ridge, stepwise regression, and generalized boosting regression) to
assign weighting to each factor dynamically, where our algorithms select the stocks with the highest
predictive ratings based off the factor that has the highest predictive return over the past training
period. Then, our algorithms employ a statistical method known as minimum variance optimization
to complete the portfolio recommendations based on the information each individual client has
provided us through the questionnaire. Minimum variance performs diversification by constraining
mean, volatility and correlation inputs to reduce sampling error. By using minimum variance, we
ensure each client has proper diversification and portfolio risk. (See “A. Methods of Analysis and
Investment Strategies” of Item 8 below.)
Stocks in a Proposed Portfolio will consist solely of those on the Stock List. The
Proposed Portfolio is made available to all of our clients through the App Platform. The client has
the sole discretion to make the ultimate decision regarding whether to invest in any Proposed
Portfolio. Our algorithms generate an updated Proposed Portfolio on the first day of each calendar
quarter for each client based on the information such client has provided, new financial data, and
other factors. Our App Platform alerts clients of the changes that our algorithms have
recommended, both in proposed dollar amount changes and proposed percentage weighting
changes. Clients may accept or decline the updated Proposed Portfolio in their sole discretion.
Individual Stock Recommendations. Our App Platform provides to clients who
subscribed for our premium service ongoing recommendations on individual stock selections. Our
algorithms will analyze the potential investment value of the stocks in the Stock List and generate
a list of twenty stocks with the most potential investment value and twenty stocks with the least
potential investment value (the “Focus Stocks”). Our algorithms determine the list of the Focus
Stocks by choosing the 20 stocks with the highest and lowest ranking of predicted returns generated
by the best performing algorithm over the past training periods. (See “A Methods of Analysis and
Investment Strategies” of Item 8 below.) Koin’s algorithms will generate such Focus Stocks once
per trading day and at the beginning of each trading day. The client has complete discretion with
respect to whether it would purchase or sell any Focus Stocks.
Our algorithms monitor client accounts. Each client will have the option to enable
notifications when the daily price movement in such client’s portfolio exceeds a certain percentage.
Each client is given the option to choose a threshold (1%, 2%, 3%, 4%, or 5%+) for daily portfolio
percentage movement that triggers notifications or choose to receive no notifications at all. If a
client chooses to receive price movement notification, when a threshold for price movement
notification is triggered, our App Platform will alert the client. Please refer to Item 13 (Review of
Accounts) on page 9 for details on client accounts review. For risks associated with not receiving
the daily price movement, please refer to B (Risk of Losses) of Item 8 on page 7 below.
We may provide additional algorithms-driven advisory services in the future.