AdaptAsset

Financial Solutions Realized

Automatic Investment Management

A Modern Perspective

WWW.ADAPTASSET.COM
Table

Who We Are 3

Objective 4

Background 5

Automatic Investment Management 6

A Modern Perspective 10

Results & Analysis 16

Disclaimer 18

ADAPTASSET 2
Who We Are
AdaptAsset is a quantitative, systematic investment firm focused on enhancing
proven strategies and methods. Backed by a wealth of history and data, our
approach is to provide innovative strategies that can excel in a variety of
environments.

The Team
Through a broad diversity of experience and knowledge, the AdaptAsset teams
brings rigor and creativity to pioneer its solutions. With expertise from research to
software engineering and risk assurance, AdaptAsset aims to harness the unique
strengths of its team members to provide an unparalleled experience for its clients.

Harrison Tateosian Jacob Mikesell Siddhartha Konda

Investment Research Software Engineering Risk Assurance

U.C. Irvine, B.A. Business U.C. Irvine, Honors B.S. Physics U.C. Irvine, B.S. Business
Economics Information Management

ADAPTASSET 3
Objective

We aim to take a closer look at Automatic Investment Management (AIM), an

investing method pioneered in the 1980’s. While it proved versatile for its time,
advances today have led our testing and analysis to be much more rigorous and
allow for significant improvements to be made over the core AIM framework.

By adapting and modifying the core AIM framework through several different
economic cycles and volatility levels, we can create an investment system that
contains many of the same elements that made AIM so popular many years ago,
while being much more agile and responsive.

ADAPTASSET 4
Background

Automatic Investment Management, or AIM for short, was an investment strategy

pioneered by Robert Lichello in 1977. It attempted to create a more disciplined and
advantageous system than traditional buy-and-hold that was in nature long-term
thinking but could also react to volatile events by mathematically deciding when to
purchase and sell securities.

AIM attempted to answer four main investment questions:

i. When starting a new investment strategy, what is the proper starting amount?

ii. How much should initially be kept in a cash reserve?

iii. When, and how much profit should be taken?

iv. When, and how much reinvestment should be implemented?

At its core, AIM was a method to increase one’s investment in a position, without
needing to put in additional follow-on capital. This can be similarly thought of as
“trading around a core position”. AIM attempts to buy low and sell high, all the
while steadily reinvesting those profits to increase the position’s size, without the
need of additional outside capital.

ADAPTASSET 5
Automatic Investment Management
The first test of any new investment strategy should be to understand its goals, and
upon achieving that, figure out how to accomplish those goals. When examining
Automatic Investment Management, we must look at it as a replacement to the
traditional investing method of buy and hold.

Trading around a core position as well as other variations are merely modifications
upon this base concept. AIM attempts to solve both functions, however, in order to
establish a starting point, we must try to analyze it purely from a buy and hold
perspective. To achieve this, we will try to isolate purely any alpha generated by the
investment conditions beyond buy and hold.

The starting parameters will ideally be as close as possible, as well as testing over
several time periods to try and eliminate any rogue effects. Recall that one of AIM’s
primary functions is to solve how much an investment should start with. To test its
similarity to buy and hold, we will start with the same conditions as buy and hold,
100% initial investment. Our goal is to see if AIM is given the same starting
parameters, whether it can generate any outperformance, or will any potential
outperformance come largely from the starting conditions and not AIM itself
[Exhibit 1]. Note that all risk statistics assume end-of-month values.

ADAPTASSET 6
Exhibit 1: 100% Initial Investment with 10% Buy/Sell Parameters (2002-2018)

Source: Quantopian, AdaptAsset. Simulated Performance

100% Invested at Inception

Compound Return 5.9%
Volatility 17.0%
Sharpe Ratio 0.42
Maximum Drawdown -51.3%
Alpha 0.57
Growth of $1 $2.61
Before we begin to modify the underlying Automatic Investment Management
framework, it’s important to understand the chart and data table. The chart shows
the growth of $1 invested from January 2002 through October 2018, where it has
grown to $2.61. This equates to an annualized compounded growth of 5.9%.

The chart helps to visualize the path undertaken by the portfolio, which includes the
approximate 50% drawdown during 2008. This is confirmed by glancing at the max
drawdown of 51.3% from peak to trough.

The portfolio volatility of 17.0% led to a Sharpe ratio of 0.42. While the Sharpe ratio
indicates that the portfolio’s return was greater than the risk-free rate, the return
must be compared to the underlying benchmark, the S&P-500. When compared to
the benchmark, the Alpha value of -0.03 indicates that the Portfolio performed

ADAPTASSET 7
worse than the Benchmark. In other words, a potential investor would have been
better off buy buying and holding the S&P-500 than following the investment
strategy.

While disappointing, there are a few possible explanations for the low Alpha value.
The portfolio was fully invested from the inception date. AIM relies on having
available cash to average down, allowing the strategy to then begin to perform as
intended. This led to the strategy only being able to raise cash from 2014 on, well
past the point when having spare cash available would have been useful to help
returns. In addition, raising cash as the market increased, while taking profits off the
table, dragged returns as the equity component of the strategy began to shrink.

Another possible condition could have risen from the fixed Buy and Sell parameter.
Having fixed Buy and Sell parameters reduces the strategy’s effectiveness to adapt
to different market environments, an effect that will be examined in a later strategy.
In addition, the magnitude of the Buy and Sell percentages are important. If the
system is not able to generate significant differences in market value from the
underlying Portfolio Control, then the strategy will not execute an order to buy or
sell. Using a diversified instrument, such as the S&P 500, may not have the volatility
level to satisfy the Buy and Sell Parameters on a regular basis.

Our goal is to exhibit scenarios under which Automatic Investment Management

can outperform its underlying benchmark by creating conditions that are more
suitable for a trading environment compared to buy and hold. We express this logic
in Exhibit 2, where we will lower the initial investment amount, lower the Buy and
Sell parameters, and increase the volatility of the instrument.

Exhibit 2: 50% Initial Investment, 5% Buy/Sell Parameters, Volatility (2002-2018)

ADAPTASSET 8
 Source: Quantopian, AdaptAsset. Simulated Performance

50% Initial Investment

Compound Return 8.2%
Volatility 19.1%
Sharpe Ratio 0.51
Maximum Drawdown -49.4%
Sortino Ratio 0.74
Growth of $1 $3.76
In Exhibit 2, the underlying benchmark was changed from the S&P-500 to the
Nasdaq 100. This increased the volatility of the portfolio from 17.0% to 19.1%, but
also led to an increase in CAGR to 8.2%. Ultimately, this resulted in an increase in
the Sharpe Ratio from 0.42 to 0.51.

For the first half of the simulation, the AIM strategy outperformed the benchmark.
However, this was likely due to the initial starting allocation of 50%. This allowed
AIM to average in at lower prices and be able to benefit from an increase in returns
over the following years. As the strategy moved to post-2008, performance of the
strategy relative to the benchmark declined, ultimately leading to the strategy
underperforming from 2016 on. This was likely due to the accumulation of cash in
the strategy as it waited for an opportunity to repurchase at a lower price.

ADAPTASSET 9
A Modern Perspective
There are several parameters that can be introduced to Automatic Investment
Management to improve performance. These improvements will center around
creating a more dynamic buy and sell order flow, allowing the strategy to use
leverage, as well as including broad market trading restrictions. The goal of these
improvements is to reduce the total drawdown, increase the flexibility of the
strategy to adapt to different volatility environments, as well as being able to
maximize the return during periods when the strategy is heavily allocated.

In Exhibit 3, the underlying benchmark was changed to 2x leveraged S&P-500. This

change increased the volatility of the strategy, allowing for a greater period to
exhibit excess returns, while preserving a long period of time to backtest the
strategy across different volatility environments.

In addition, a portfolio rebalance was added to the investment strategy that will
reinvest extra cash above a certain threshold at the beginning of each year. This
prevents the strategy from accumulating an excess amount of cash during long
periods of strong performance in the underlying security, while still preserving free
capital to take advantage of any drawdowns that could occur.

ADAPTASSET 10
Exhibit 3: 5% Buy/Sell Parameters, Rebalance, 2X Index (2006-2018)

Source: Quantopian, AdaptAsset. Simulated Performance

Rebalance
Compound Return 9.0%
Volatility 36.2%
Sharpe Ratio 0.42
Maximum Drawdown -85.0%

Sortino Ratio 0.59

Growth of $1 $2.89
Compared to Exhibit 2, the compound return in Exhibit 3 was increased from 8.2%
to 9.0%, a significant improvement. However, the higher leverage led to a higher
level of volatility of 36.2%, as well as a higher drawdown of 85%. These impacted
the risk-adjusted returns of the strategy by lowering the Sharpe Ratio to 0.42 and
the Sortino Ratio from 0.74 to 0.59.

The impact from rebalancing was likely minimal as the only time that the rebalance
would have taken effect was during the final years of the backtest. Given that most
of the volatility occurred in the first half of the simulation, the backtest likely had
little effect on the strategy’s overall performance.

ADAPTASSET 11
In Exhibit 4, we will continue to run the strategy based on the 2X leveraged S&P-
500 and focus on controlling the excess risk. To accomplish this, the strategy will
use a set of underlying moving averages to dictate when the strategy can and
cannot trade. The goal of the moving average implementation is to prevent the
strategy from being active during large downtrends in the market. This will hurt the
strategy’s ability to buy low during drawdowns in the underlying security but will be
made up for by lower overall losses, thereby reducing the strategy’s overall risk
parameters. The moving average parameter will function such that when the
average is declining, the strategy will hold only cash, and when the average inflects
and begins to rise, the strategy will reinvest most of the portfolio.

Exhibit 4: 5% Buy/Sell Parameter, Rebalance, Moving Average (2006-2018)

Source: Quantopian, AdaptAsset. Simulated Performance

Moving Average
Compound Return 13.6%
Volatility 22.7%
Sharpe Ratio 0.68
Maximum Drawdown -42.7%

Sortino Ratio 0.95

Growth of $1 $4.79

ADAPTASSET 12
Exhibit 4 shows a substantial improvement in performance over Exhibit 3. The
compound return has increased from 9.0% to 13.6% alongside a reduction in
volatility from 36.2% to 22.7%. This led to a large increase in risk-adjusted return as
the Sharpe Ratio increased to 0.68 and the Sortino Ratio increased to 0.95.

This large improvement in performance was likely due to the conditional periods
that the Moving Average parameter allowed the strategy to trade. Looking at the
performance graph, we can see that the Moving Average parameter allowed the
system to avoid most of the drawdown during the 2008 Recession. This helped to
reduce the drawdown to 42.7% from 85.0% previously. In addition, the added
parameter had a limited effect on capping upside return as there were few periods
where the strategy held cash for extended periods of time outside of the recession.

In Exhibit 5, we will attempt to improve the strategy’s responsiveness by adjusting

the buy and sell parameters. Currently, the strategy uses a fixed percentage system
which triggers an order any time that the difference between the current value and
the portfolio control exceeds the buy/sell threshold. While this allows for a simpler
calculation, it forces the strategy to be less responsive in the face of different
volatility environments. By shifting the buy/sell parameter towards one based on
the current volatility level, this will allow the strategy to become more responsive to
the local volatility environment. During high volatility events, such as a significant
drawdown, having an adaptable buy/sell system will respond better than a fixed
strategy by having orders placed at wider scales rather than at fixed intervals.
Conversely, during lower volatility periods, an adaptive volatility measurement will
allow the strategy to trade at smaller intervals compared to the fixed strategy.
Exhibit 5 will utilize the same parameters as Exhibit 4 except for a changed buy/sell
parameter.

ADAPTASSET 13
Exhibit 5: Adaptive Buy/Sell, Rebalance, Moving Average (2006-2018)

Adaptive Buy/Sell
Compound Return 14.1%
Volatility 23.0%
Sharpe Ratio 0.69
Maximum Drawdown -42.8%
Sortino Ratio 0.97
Growth of $1 $5.06
Exhibit 5 showed an increase in compound return from 13.6% to 14.1%. This was
achieved with almost the same volatility profile, which helped to increase the
Sharpe Ratio and Sortino Ratio to 0.69 and 0.97, respectively. While the adaptive
buy/sell proved beneficial, the effect was muted as there were several periods
where the strategy exhausted its cash reserve, hampering the flexibility of the
adaptive order parameter.

In Exhibit 6, we will attempt to improve performance by allowing the strategy to

borrow additional assets and exceed a 1.0 leverage ratio. The goal will be to provide
additional returns during periods when the strategy is performing well, and to rely
on the moving average parameter to avoid excess losses during drawdowns. In
addition, allowing borrowing will prevent periods when the strategy cannot perform
optimally due to insignificant cash holdings.

ADAPTASSET 14
Exhibit 6: Adaptive Buy/Sell, Moving Average, Leverage

Leverage
Compound Return 15.5%
Volatility 24.3%
Sharpe Ratio 0.72
Maximum Drawdown -42.9%
Sortino Ratio 1.02
Growth of $1 $5.91
Exhibit 6 shows a significant improvement in compound return from 14.1% to 15.5%,
along with a modest increase in volatility from 23.0% to 24.3%. The improved
performance helped to increase the risk-adjusted metrics of Sharpe and Sortino
Ratios to 0.72 and 1.02 respectively.

The increase in available leverage allowed the strategy to take advantage of higher
volatility events where previously Automatic Investment Management would have
exhausted its cash reserve before completing all its purchase orders. The Moving
Average parameter continued to provide downside protection, allowing the
leverage to boost returns to the upside without significantly increasing the risk
profile.

ADAPTASSET 15
Results & Analysis
Automatic Investment Management is a strategy centered around building a long-
term position without the need for additional capital. The strategy focuses on
growing portfolio control, a metric that measures the amount of total capital
invested into a position. The strategy uses buy and sell parameters to trade around
the core position by buying low and selling high to steadily grow the amount of
invested capital in a position.

While promising in concept, the strategy has a hard time outperforming its
underlying benchmark. As seen in Exhibit 1, the strategy suffers from a lack of
available capital to use during drawdowns, as well as tending to sell excess amounts
during periods of positive returns. This culminates in the strategy selling too early,
and not having enough free capital to take advantage of potential drawdowns.

Increasing the underlying volatility helps the strategy to perform better as a higher
volatility level presents more opportunities for the strategy to boost returns by
trading around the core position. However, as seen in Exhibit 2, while this is
beneficial, the strategy is much more influenced by the starting allocation, an
arbitrary variable that masks the ability of the underlying strategy to perform.

We examined several different methods to increase returns of the strategy while still
preserving its core framework. In Exhibit 3, the underlying benchmark was increased
to a leveraged fund to increase the volatility to help the strategy perform better, as
well as introducing an annual rebalance to ensure that the strategy does not
accumulate too much excess cash for extended periods of time.

Increasing the volatility of the underlying benchmark led to a higher volatility

exhibited by the strategy, resulting in the next variation, implementation of a
moving average parameter, that attempted to signal periods of time when the

ADAPTASSET 16
strategy could trade and when it should hold cash instead. Exhibit 4 showed that
implementation of a moving average strategy proved significantly beneficial to the
strategy as it allowed the strategy to avoid several periods of significant drawdowns
without hampering the upside performance.

Exhibit 5 replaced the traditional buy/sell method of Automatic Investment

Management with a variable system that allowed the strategy to adapt to changing
volatility levels, rather than using a fixed measurement for the entire life of the
strategy. This would allow the strategy to use wider buy/sell parameters during
higher volatility periods and closer scales during low volatility environments.
Modifying the buy/sell logic helped to improve performance, but the strategy was
still hampered by a lack of available capital during drawdown periods, reducing
performance.

Additional leverage was added in Exhibit 6, allowing the strategy to use extra capital
when it traditionally would have exhausted its cash reserve. This allowed the
strategy to take advantage of drawdown periods by buying more than it otherwise
would have been able to without dramatically increasing the risk profile.

Through several targeted modifications, we can significantly improve the risk-

adjusted performance of Automatic Investment Management.

ADAPTASSET 17
Disclaimer
PAST PERFORMANCE IS NOT NECESSARILY INDICATIVE OF FUTURE RESULTS

Forward-Looking Information. This presentation may contain “forward-looking information”. Because such
forward-looking information involves risks and uncertainties, actual results of the funds or accounts may
differ materially from any expectations, projections or predictions made or implicated in such forward-
looking information. Prospective investors are therefore cautioned not to place undue reliance on such
forward-looking statements. In addition, in considering any prior performance information contained in
this presentation, prospective investors should bear in mind that past results are not necessarily indicative
of future results, and there can be no assurance that the funds or any account will achieve results
comparable to those discussed in this presentation. This presentation speaks as of the date hereof and
neither AdaptAsset nor any affiliate or representative thereof assumes any obligation to provide any
recipient of this presentation with subsequent revisions or updates to any historical or forward-looking
information contained in this presentation to reflect the occurrence of events and/or changes in
circumstances after the date hereof.

General information regarding hypothetical performance and simulated results. These results are based on
simulated or hypothetical performance results that have certain inherent limitations. Unlike the results in
an actual performance record, these results do not represent actual trading. Also, because these trades
have not actually been executed, these results may have under- or over-compensated for the impact, if
any, of certain market factors, such as lack of liquidity. Simulated or hypothetical trading programs in
general are also subject to the fact that they are designed with the benefit of hindsight. No representation
is being made that any account or fund will or is likely to achieve profits or losses similar to those being
shown. The results do not include other costs of managing a portfolio (such as custodial fees, legal,
auditing, administrative or other professional fees). The information in this presentation has not been
reviewed or audited by an independent accountant or other independent testing firm. More detailed
information regarding the way the charts were calculated is available on request. Any actual fund or
account that AdaptAsset manages will invest in different economic conditions, during periods with
different volatility and in different securities than those incorporated in the hypothetical performance
charts shown. There is no representation that any fund or account will perform as the hypothetical or
other performance charts indicate.

© 2018 AdaptAsset. All rights reserved.

ADAPTASSET 18
ADAPTASSET 19