Alternative Risk Premia

October 2019

Guillaume Monarcha | PhD, Head of Research

Orion ARP Indices – October 2019 Summary

▪ In October, the ARP industry posted an average return of

-0.1%, bringing the year-to-date performance to a
Equity - Value 0.85%
mitigated +0.2%.
Equity - Quatity -1.00%
▪ In terms of asset classes, commodity strategies posted the
Equity - Low risk -0.97%
best average return (+ 0.7%), while interest rates and FX
Equity - Momentum -1.16%
ARPs averaged -0.5%. In the middle of the table, equity
Equity - Trend 0.26%
strategies ended the month with a flat performance.
Equity - Mean-rev. 0.68%
▪ Within the equity space, academic strategies
Equity - Short vol. 0.33%
underperformed trading ones, with the notable exception
Commodity - Carry -0.52%
of value strategies that recorded their second positive
Commodity - Carry (trd) 0.88%
monthly performance in 9 months. We note that quality
Commodity - Liquidity 0.35%
ARPs recorded one of their worst performance (below 5th
Commodity - Trend -1.11%
percentile), for the third time of the year.
Commodity - Short vol. 2.81%
▪ Commodity strategies posted heterogenous
Interest rates - Carry -0.50%
performances, from -1.1% for trend ARPs to +2.8% for
Interest rates - Momentum -1.89%
short volatility (top 5% monthly return).
Interest rates - Trend -0.63%
▪ In the same vain, the monthly returns of FX strategies
Interest rates - Short vol. -0.24% ranged from -2.4% (trend) to +0.7% (short volatility).
FX - Carry -0.68% ▪ Finally, interest rate strategies globally suffered, with
FX - Value 0.26% average returns between -1.9% for momentum and -0.2%
FX - Mean-rev. 0.33% for short volatility.
FX - Short vol. 0.66%

FX - Trend -2.39%

-5 .0% -2 .5% 0.0% 2.5% 5.0%

FOCUS | ARP performances: is there any alpha left?

Source: Orion Financial Partners

 Focus | ARP performances: is there any alpha left?
With a year-to-date return of +0.2% at the end of decisions are driven by the historical performance of
October, the recent performance of the alternative risk the investable indices provided by investment banks,
premia (ARP) industry hits record low. This figure is all which rarely go back before the early 2000s. We
the more worrisome that the 4 worst yearly therefore understand why the current performance
performances (including 2019) were recorded within levels are so problematic.
the last 5 years (chart 1).
Several factors can explain this recent deterioration
Chart 1: Yearly performances of the ARP industry (Orion in the performances of ARPs. The first one is the
Global ARP index) backtesting bias embedded in the historical
10%
performances provided by ARP providers. This bias is
inherent to the incorporation of the backtested
8%
performance – which is theoretical – in the historical
6%
data of ARP indices before their launch date. After the
4% effective commercialization of the index, the
2% performance displayed is ‘live’ or real performance. On
0%
average, Naya and Tuchschmid (2018) estimate that
2005 2007 2009 2011 2013 2015 2017 2019 (YtD) the Sharpe ratios of ARP strategies decrease by 75% on
-2%
average once they are effectively commercialized. In a
-4%
previous study1, we found more mitigated results,
Source: Orion Financial Partners
suggesting that half of that discount should be
From a very long-term perspective, one can check attributed to conjunctural or structural tightening of
that current performance levels are close to those the various risk premia. A second explanation could be
recorded before the mid-70s (chart 2), despite being linked to the sensitivity of certain ARP strategies to the
significantly lower than during the 1975-2008 period. global macro environment. In line with Blin, Ielpo, Lee,
and Teiletche (2018), Varsani and Jain (2018), or
Chart 2: Average 5 year rolling annualized performances
Bellone, Declerck, Nordine, and Vy (2019), we have
of academic factors (multi-asset, multi-strategy)
previously shown2 that half of the ARP strategies
8%
exhibit significant exposures to risk aversion, liquidity
7% risk, economic growth, and inflation. A third potential
6% factor that could explain this deterioration in the
5% average performance level is the tremendous amount
4% of liquidity that has been injected into the financial
3% system since the global financial crisis. This large
2% liquidity flow could have led to ‘blind’ (or less selective)
1% investment decisions, leading to a significant
0% divergence between asset prices and their
1925 1935 1945 1955 1965 1975 1985 1995 2005 2015
fundamentals. Under this assumption, the
Source: Ilmanen, Israel, Moskowitz, Thapar & Wang (2019), performances of the various ARP strategies should
available at www.aqr.com
recover gradually, along with the forthcoming market
However, one will agree that current investor corrections. A last explanation – which is more
concerns are quite rightly driven by this ‘relatively structural and therefore more worrisome – could be
recent’ history. Indeed, while the central argument of the impact of the strong development of the industry
ARP providers is based on the long-term performance during the last decade. Indeed, one can assume that
of the various academic risk premia, investors' the strong asset inflows experienced by certain

1 Orion Financial Partners (2019), “Alternative Risk Premia”, May. 2 Orion Financial Partners (2019), “Alternative Risk Premia”, September.

Alternative Risk Premia | October 2019 Orion Financial Partners | p.2

strategies led to a progressive deterioration in the risk
premia they are designed to capture, therefore leading
to a structural and potentially durable deterioration in
their performances.
Quantitative analysis of ARP performances
To bring a first piece of answer to the question of
the deterioration in the average return level of ARPs,
we propose to decompose their performance into two
elements: a conjunctural performance driver (linked
with the macroeconomic environment) and an
absolute performance driver (the ‘alpha’ carried by the
various risk premia). In our analysis, we do not
incorporate any market-based performance drivers as
ARPs are alternative investment strategies that, by
construction3, exhibit low exposure to market risks.
The data
ARP data. For the various ARP strategies, we
consider the monthly returns of Orion ARP indices.
These indices are built as equal-weighted average
portfolios of individual ARP indices provided by 10
investment banks4. These indices differ from academic
factors at several levels: they are investable, they
incorporate transaction costs, and they encompass
both academic and trading risk premia.
Macroeconomic environment. Following Ilmanen
(2011) and Ilmanen, Maloney, and Ross (2014), we
characterize the global market environment around
four dimensions: economic cycle, illiquidity, global risk
aversion, and inflation. We proxy the economic cycle
by our live Global Macroeconomic Cycle Index for
developed countries, a nowcasting index based on the
analysis of macroeconomic statistics from 11
countries5 (see Monarcha and Le Ny, 2017). For
inflation, we consider the US 10 years inflation
breakeven (proxy of inflation expectations), i.e. the
yield difference between 10 years treasuries and 10
years US inflation linked bonds. Risk aversion is proxied
by the weighted average of percentile ranks of the
3 This is especially the case for academic ARPs that are built as long/short
beta neutral portfolios.
4
See Orion Financial Partners, « ARP monitor – March 2019 ».
5 USA, Germany, France, Italy, Spain, UK, Switzerland, Japan, Australia,
Canada, and Korea.
6 This methodology in similar to those used by investment banks to
construct specific risk appetite indices, see Illing and Aaron (2005) or Gex
and Coudert (2006) for a review of risk appetite indices.
Alternative Risk Premia | October 2019
European and US equity implied volatilities (VIX and
V2X indices), and of the US and Euro high yield
spreads6. Finally, we proxy liquidity risk by a similar
index, based on the TED spreads and the swap spreads
for the Eurozone and the US. To ease interpretation,
the four macroeconomic factors are standardized over
the estimation sample.
The model
For each ARP index, we estimate the following
dynamic multifactor model:
𝑅𝑖,𝑡 = 𝛼𝑖,𝑡 + ∑ 𝛾𝑖,𝑗,𝑡 𝐹𝑗,𝑡 + 𝜀𝑖,𝑡
𝑗
𝛼𝑖,𝑡 = 𝛼𝑖,𝑡−1 + 𝑢𝑖,𝑡
𝛾𝑖,𝑗,𝑡 = 𝛾𝑖,𝑗,𝑡−1 + 𝑣𝑖,𝑗,𝑡
where 𝑅𝑖,𝑡 are the monthly rescaled returns (5% full-
sample volatility) of ARP index 𝑖, 𝛼𝑖,𝑡 the absolute
performance carried by the premia, 𝐹𝑗,𝑡 the
standardized intra-month average value of factor 𝑗,
𝛾𝑖,𝑗,𝑡 the sensitivity of ARP 𝑖 to factor 𝑗. As our aim is to
capture structural evolutions in both 𝛼𝑖,𝑡 and 𝛾𝑖,𝑗,𝑡 , we
control for their dynamics7.
Therefore, as 𝐹𝑗,𝑡 are standardized over the
estimation sample, ∑𝑗 𝛾𝑖,𝑗,𝑡 𝐹𝑗,𝑡 measures the
conjunctural deviations generated by the sensitivity of
the various ARPs to the macroeconomic environment.
𝛼𝑖,𝑡 can then be viewed as the time varying expected
absolute return, after controlling for conjunctural
variations.
Estimated alphas
Charts 3 to 7 illustrate the evolution of the
estimated annualized absolute performances for the
various ARP strategies. We summarize the main results
are the following:
▪ The average absolute performance generated by
ARP strategies has been almost divided by 2
7 𝑣𝑎𝑟(𝑢𝑖 ) = 𝑣𝑎𝑟(𝑣𝑖,𝑗 ) = 0.005𝑣𝑎𝑟(𝜀𝑖 ). the model is estimated using the
Kalman Filter, through a custom modified version of the EM algorithm
incorporating a gate to control for the impact extreme signal observations
on the state variables.
Orion Financial Partners | p.3
 since 2010, dropping from 7% to 3.5% at the end ▪ Interest rate ARPs stand out from other
of September 2019. This effect prevails for both strategies. As illustrated on chart 6, their
academic and trading ARP strategies. average level of absolute performance remained
unchanged for carry and short volatility, and
▪ Equity strategies exhibit the most significant even increased for trend and momentum.
deterioration in alpha (-65% on average), from -
Chart 5: Time-varying absolute performance delivered by
98% for the value premia (from 13% to 0%) to a
commodity ARP strategies (annualized)
contained -18% for quality strategies (chart 4).
18%
The statement is the similar for commodity 16%

strategies, except for short volatility ARPs whose 14%

12%
alpha remained almost stable (chart 5). 10%
8%
Chart 3: Time-varying absolute performance delivered by 6%

academic and trading ARP strategies (annualized) 4%

2%
8%
0%
7% 2000 2005 2010 2015

6% Carry Carry tr. Liquidity

Trend Short volatility
5%
Source: Orion Financial Partners
4%

3% Chart 6: Time-varying absolute performance delivered by

2% interest rate ARP strategies (annualized)
1%
10%
0%
2000 2005 2010 2015 8%

All Trading Academic 6%

Source: Orion Financial Partners
4%

Chart 4: Time-varying absolute performance delivered by 2%

equity ARP strategies (annualized) 0%

2000 2005 2010 2015
20%
-2%
Carry Momentum
15%
Trend Short Volatility
10% Source: Orion Financial Partners

5%
Chart 7: Time-varying absolute performance delivered by
FX ARP strategies (annualized)
0%
2000 2005 2010 2015 10%

-5%
8%
Value Quality Low Risk
Momentum Trend Mean Reversion 6%
Short Volatility
4%

Source: Orion Financial Partners 2%

0%
▪ Within FX ARPs, the alpha of trend and carry 2000 2005 2010 2015
-2%
strategies show significant decrease. Conversely, Value Mean Reversion
the absolute component of value, short Short Volatility Trend
Carry
volatility, and mean reversion risk premia
remained stable since 2000 (chart 7). Source: Orion Financial Partners

Alternative Risk Premia | October 2019 Orion Financial Partners | p.4

 We identify 3 potential explanation to this global
decrease in the absolute performance carried by ARP
strategies. The first one is the effect of backtesting bias.
It stems from the propensity of ARP providers to over-
optimize their models during the backtesting process.
Consequently, as the track record of ARP indices
combine the backtested (or pre-launch) and live
performances (post-launch), one can argue that the
average performance level before the effective
commercialization of the strategy is overestimated.
Naya and Tuchschmid (2019) estimate that this effect
induces an average deterioration of 75% in the Sharpe
ratios. In a previous study8, we find a more mitigated
effect. However, under this hypothesis, absolute
performance carried by academic risk factors should
not be impacted. But as shown on chart 8, we observe
the same effect when estimating the time-varying
alpha carried in academic factors9, which lead us to
reject this (isolated) explanation.
Chart 8: Time-varying annualized alpha carried by ARP
strategies and academic factors (rhs, annualized)
8%
7%
6%
5%
4%
3%
2%
1%
0%
2000 2005 2010 2015
ARPs Academic factors
Source: Orion Financial Partners
A second explanation could be that the tremendous
amount of liquidity injected by the central banks in the
financial system since the global financial crisis led to
“blind” asset allocation patterns, thus implying a
progressive disconnection between asset prices and
their fundamental drivers. However, such an effect
should have been impacting almost homogenously all
the market segments including interest rates, which is
not the case (chart 6). The last potential explanation is
linked to the strong development ARP industry10 itself
8 Orion Financial Partners (2019), “Alternative Risk Premia”, May.
9 We estimated our model over the same sample (2000-2019), from a set
of 44 multi-asset / multifactor factors, taken from: Ilmanen, Israel,
Moskowitz, Thapar & Wang (2019), available at www.aqr.com.
10 And more broadly of the factor industry.
Alternative Risk Premia | October 2019
during the last decade, that could have led to the
natural disappearance of certain risk premia.
Summary
The recent deterioration in the absolute
performance carried by the various risk premia is a
major concern for ARP investors. Our analysis show
that, at the industry level, it has been almost divided by
since 2010, to reach 3.5%11 at the end of September
2019 (chart 2). From here, should investors turn away
from alternative risk premia, and more generally factor
investing? We do not believe so, for several reasons.
First, whereas the average level of alpha has
significantly decreased, we observe heterogenous
pattern across the various investment strategies, and
across the various asset classes.
Second, as recently mentioned in both academic
and portioners researches, certain risk premia exhibit
significant sensitivity to the market environment. This
is confirmed by our analysis. On average, found that
5%
ARP strategies exhibit significant negative sensitivity to
4% liquidity risk (negative) and risk aversion (positive,
chart 9). Once again, as discussed in our previous ARP
3%
monitor12, the sensitivities of the various ARP
2%
strategies to the various macroeconomic factors may
1% differ significantly. Consequently, as shown on chart
0%
10, a significant part of ARP performance stems from
exposures to the macroeconomic factors. This
conjunctural component of their expected returns
tend to deliver higher performances during
crisis/recession episodes than during growth/steady
ones, and account for 0% to 50% of the total expected
returns (chart 11).
What implications for investors?
The main conclusion we can draw from this analysis
is that investing in ARPs is no more a “free lunch”.
Indeed, the current practice that consists in investing
in well diversified ARP portfolios13, and that have
shown significant results since the mid-70s, appears to
be highly under-optimal today. Given the significant
changes in the expected returns of ARPs since 2010,
11 Annualized, for a 5% volatility level over the full sample.
12 Orion Financial Partners, “ARP monitor – September 2019”.
13 Most ARP portfolios are allocated following risk parity, or risk-based
allocation techniques.
Orion Financial Partners | p.5
one can no more rely on their long term historical Chart 11: Relative importance of conjunctural
statistical properties. Instead of turning away from the performance and alpha in monthly expected returns
100%
ARP industry, investors should consider sound 90%

investment process that rely on the quantitative 80%

70%
selection of ARPs, according to: 60%
50%

– the current level of alphas 40%

30%
– their sensitivity to the global macro 20%

environment, either with a tactical allocation 10%

0%
perspective (i.e. with a focus to capture the 2000 2005 2010 2015

conjunctural performance of ARPs), either with Cunjunctural performance Alpha

the underlying objective of achieving a target Source: Orion Financial Partners

exposure to a specific macroeconomic factor. Source: Orion Financial Partners

References
Chart 9: Average time varying exposures of alternative risk
premia to growth, inflation, risk aversion and liquidity risk
Bellone, Declerck, Nordine, Vy (2019), “Multi Asset
0.40%
Style Factors Have Their Shining Moments”, working
0.30%
paper.
0.20%

0.10% Blin, Ielpo, Lee, Teiletche (2018), “Factor Timing

0.00%
Revisited, Alternative Risk Premia Allocation Based on
2000 2005 2010 2015
-0.10%
Nowcasting and Valuation Signals”, working paper.
-0.20%
Liquidity risk Risk aversion Gex and Coudert (2006), “Can risk aversion indicators
Economic cycle Inlfation expectations anticipate financial crises?”, Banque de France,
Source: Orion Financial Partners Financial Stability Review.

Chart 10: Average alpha and “conjunctural performance” lling and Aaron (2005), “A Brief Survey of Risk-
of ARP strategies Appetite Indexes”, Bank of Canada, Financial System
12%
Review,
10%

8%
Ilmanen (2011), “Expected returns”, book, Wiley Ed.
6%

4%
Ilmanen, Maloney, and Ross (2014), “Exploring
2%
Macroeconomic Sensitivities: How Investments
0%

-2%
2000 2005 2010 2015 Respond to Different Economic Environments”, Journal
-4% of Portfolio Management, vol. 40(3).
α(t) ∑γ(t)F(t) α(t)+∑γ(t)F(t)
Monarcha, Le Ny (2017), “Quantitative Decryption of
Source: Orion Financial Partners the Market Environment. Part 1 – The
Macroeconomic Cycle”, working paper, Orion
Financial Partners.

Naya and Tuchschmid (2019), “Alternative Risk

Premia: Is the Selection Process Important?”, The
Journal of Wealth Management, Summer 2019, 22 (1)
25-38.

Alternative Risk Premia | October 2019 Orion Financial Partners | p.6

 Equity Risk Premia

Performance and risk metrics | October 2019

Number of
Monthly Percentile Information Maximum Calmar Current
underlying YtD Volatility2
1
performance rank ratio3 drawdown ratio4 drawdown
risk premia
Academic risk premia
Value 19 0.8% 85.4% -5.5% 5.1% -0.09 -14.3% -0.03 -11.5%
Europe 7 1.0% 84.3% -7.0% 5.1% -0.08 -12.7% -0.03 -11.1%
USA 6 0.2% 55.6% -4.1% 4.8% -0.23 -21.7% -0.05 -18.5%
Quality 13 -1.0% 5.1% -0.5% 4.5% 0.67 -4.2% 0.72 -2.3%
USA 4 -1.4% 2.8% -1.3% 3.9% 0.21 -5.1% 0.16 -3.9%
Europe 4 -0.6% 18.0% -2.1% 4.9% 0.93 -6.0% 0.76 -6.0%
Low risk 16 -1.0% 11.8% 4.4% 4.9% 0.86 -7.8% 0.54 -1.0%
USA 4 -1.8% 5.1% 7.1% 4.4% 0.77 -9.3% 0.36 -1.8%
Europe 6 -0.7% 16.9% 2.4% 4.9% 0.94 -7.0% 0.66 -0.7%
Momentum 14 -1.2% 14.6% -3.1% 6.1% 0.35 -13.2% 0.16 -8.0%
Europe 6 -0.6% 24.2% -1.6% 6.3% 0.62 -9.7% 0.41 -6.5%
USA 3 -2.3% 7.3% -5.3% 5.7% 0.06 -19.7% 0.02 -16.1%
Trading risk premia
Trend 4 0.3% 53.4% -4.1% 5.7% 0.59 -17.0% 0.25 -12.3%
Mean-reversion 10 0.7% 64.6% 2.5% 7.2% 0.62 -7.3% 0.63 -4.4%
Short volatility 42 0.3% 36.0% 2.7% 7.4% 0.47 -10.7% 0.28 -2.8%
1
At the end of the month. 2Average volatility of individual ARPs (since 2010, EWMA, 63 trading days, λ=0.94). 3Average return / average volatility (since 2010).
4Average return / average maximum drawdown.

Source: Orion Financial Partners

Average/min/max of monthly performances Current vs. min/max historical volatility

5.0% 20%
4.0% - - - min/max Average 18% - - - 5%/95% percentiles
3.0% 16%
Historical average
2.0% 14%
1.0% 12%
Current
0.0%
10%
-1 .0%
8%
-2 .0%
-3 .0% 6%
-4 .0% 4%
-5 .0% 2%
Value Qu ality Low Risk Momentum Trend Mean Short 0%
Reversion Volatility Value Qu ality Low Risk Momentu m Trend Mean Short
Reversion Volatility
Source: Orion Financial Partners

Source: Orion Financial Partners

Average/min/max cross-sectional dispersion Average/min/max volatility

Note: average intra-index 6 months cross-sectional dispersion, annualized.

Note: the grey area represents the min/max volatility levels of the various
indices. EWMA volatility (63 trading days, λ=0.94).

Alternative Risk Premia | October 2019 Orion Financial Partners | p.7

 Commodity Risk Premia

Performance and risk metrics | October 2019

Number of
Monthly Percentile Information Maximum Calmar Current
underlying YtD Volatility2
1
performance rank ratio3 drawdown ratio4 drawdown
risk premia
Academic risk premia
Carry 8 -0.5% 22.5% -6.0% 6.5% -0.11 -7.9% 0.06 -7.9%
Trading risk premia
Carry 9 0.9% 64.6% 1.5% 3.8% 0.47 -3.1% -0.39 -0.3%
Liquidity 6 0.3% 55.1% 0.4% 3.4% -0.28 -4.4% 0.17 -3.2%
Trend 5 -1.1% 20.2% -11.5% 7.9% -0.73 -19.9% 0.24 -19.9%
Short volatility 12 2.8% 94.9% -2.3% 0.0% -0.22 -12.6% 0.09 -10.2%
1Atthe end of the month. 2Average volatility of individual ARPs (since 2010, EWMA, 63 trading days, λ=0.94). 3Average return / average volatility (since 2010).
4Average return / average maximum drawdown.

Source: Orion Financial Partners

Average/min/max of monthly performances Current vs. min/max historical volatility

8.0% 18%
16%
6.0% - - - 5%/95% percentiles
14%
Historical average
4.0% 12%
Current
2.0% 10%
8%
0.0%
6%
-2 .0% - - - min/max Average
4%
-4 .0% 2%
Carry (aca.) Carry (trading) Liquidity Trend Short Volatility
0%
Carry (aca.) Carry (trading) Liquidity Trend Short Volatility
Source: Orion Financial Partners

Source: Orion Financial Partners

Average/min/max cross-sectional dispersion Average/min/max volatility

Note: average intra-index 6 months cross-sectional dispersion, annualized. Note: the grey area represents the min/max volatility levels of the various
indices. EWMA volatility (63 trading days, λ=0.94).

Alternative Risk Premia | October 2019 Orion Financial Partners | p.8

 Interest Rate Risk Premia

Performance and risk metrics | October 2019

Number of
Monthly Percentile Information Maximum Calmar Current
underlying YtD Volatility2
1
performance rank ratio3 drawdown ratio4 drawdown
risk premia
Academic risk premia
Carry 12 -0.5% 7.3% 1.2% 7.5% 0.32 -2.2% -0.18 -0.5%
Momentum 4 -1.9% 2.2% 4.9% 2.8% 0.48 -7.1% -0.22 -3.4%
Trading risk premia
Trend 9 -0.6% 16.3% 6.9% 3.9% 0.38 -6.4% -0.19 -2.2%
Short volatility 16 -0.2% 21.3% -3.7% 4.3% -0.10 -3.8% 0.05 -3.8%
1At the end of the month. 2Average volatility of individual ARPs (since 2010, EWMA, 63 trading days, λ=0.94). 3Average return / average volatility (since 2010).
4
Average return / average maximum drawdown.
Source: Orion Financial Partners

Average/min/max of monthly performances Current vs. min/max historical volatility

1.0% 10%
- - - min/max Average 9% - - - 5%/95% percentiles
0.0%
8% Historical average
-1 .0% 7% Current
6%
-2 .0%
5%
-3 .0% 4%

-4 .0% 3%
2%
-5 .0%
1%
Carry (aca.) Momentum Trend Short Volatility
0%
Source: Orion Financial Partners Carry (aca.) Momentu m Trend Short Volatility

Source: Orion Financial Partners

Average/min/max cross-sectional dispersion Average/min/max volatility

Note: average intra-index 6 months cross-sectional dispersion, annualized. Note: the grey area represents the min/max volatility levels of the various
indices. EWMA volatility (63 trading days, λ=0.94).

Alternative Risk Premia | October 2019 Orion Financial Partners | p.9

 FX Risk Premia

Performance and risk metrics | October 2019

Number of
Monthly Percentile Information Maximum Calmar Current
underlying YtD Volatility2
1
performance rank ratio3 drawdown ratio4 drawdown
risk premia
Academic risk premia
Carry 22 -0.7% 28.7% 5.4% 5.8% 0.22 -17.8% -0.06 -2.7%
Value 5 0.3% 53.4% 4.6% 3.9% 0.73 -7.5% -0.35 -
Trading risk premia
Mean-reversion 4 0.3% 51.1% 2.7% 4.3% 0.11 -8.0% -0.04 -1.4%
Short volatility 15 0.7% 77.5% 2.6% 5.4% 1.19 -10.1% -0.22 -
Trend 8 -2.4% 4.5% -2.3% 4.0% -0.57 -11.7% 0.22 -11.7%
1At the end of the month. 2Average volatility of individual ARPs (since 2010, EWMA, 63 trading days, λ=0.94). 3Average return / average volatility (since 2010).
4Average return / average maximum drawdown.

Source: Orion Financial Partners

Average/min/max of monthly performances Current vs. min/max historical volatility

4.0% 18%

3.0% - - - min/max Average 16% - - - 5%/95% percentiles

2.0% 14% Historical average
1.0%
12%
Current
0.0%
10%
-1 .0%
-2 .0% 8%

-3 .0% 6%
-4 .0% 4%
-5 .0% 2%
Value Mean Reversion Short Volatility Trend Carry
0%
Value Mean Reversion Short Volatility Trend Carry
Source: Orion Financial Partners

Source: Orion Financial Partners

Average/min/max cross-sectional dispersion Average/min/max volatility

Note: average intra-index 6 months cross-sectional dispersion, annualized. Note: the grey area represents the min/max volatility levels of the various
indices. EWMA volatility (63 trading days, λ=0.94).

Alternative Risk Premia | October 2019 Orion Financial Partners | p.10

 Strategy definitions
These definitions are taken from G. Monarcha (2019), “An Introduction to Alternative Risk Premia”, Alternative Investment Analyst Review, Q1 2019, CAIA.

We distinguish two types of ARPs in bank’s offering. Academic ARPs are based on factors that have been well documented in the academic literature, and that
essentially involve the trading of listed and liquid products. Trading ARPs encompass a set of systematic and rule-based quantitative investment strategies, that
alternately aim to replicate hedge fund strategies (trend following, M&A…) or to exploit market anomalies, whose economic rationale may be hard to apprehend.
Unlike academic ARPs, trading ARPs are mostly backed by applied research, academic research being limited by data availability (especially in the case they rely
on the trading of OTC derivatives) or by the lack of theoretical foundations. They also differ from academic ARPs in their construction process. They are not
necessarily market neutral and may be based on a more discretionary stock selection process.

Risk premia Economic intuition Implementation Asset class References

Benefit from the price convergence between Fama et French (1992,
undervalued and overvalued assets. The Equities, rates, 1993); Asness, Frazzini
Buy undervalued securities,
Value relative value of a security is evaluated by an credit, FX, (2013); Asness,
sell overvalued securities.
economic measure (price to book ratio for commodities. Moskowitz, Pedersen
the shares, PPP for the currencies ...) (2013).
Jagadeesh et Titman
The momentum premium is based on a (1993); Carhart (1997);
Long positions in the best-
behavioral bias: demand for securities with Equities, rates, Rouwenhorst (1998);
performing stocks, short
Momentum the best recent performance tends to be credit, FX, Moskowitz and
positions in the least
larger than demand for securities with commodities. Grinblatt (1999);
performing stocks.
weaker recent performance. Asness, Moskowitz,
Pedersen (2013).
Academic risk premia

According to the CAPM theory, investors

who cannot use leverage are forced to
Ang, Hodrick, Xing,
allocate their assets in a non-optimal
Long positions on the least Equities, rates, Zhang (2006); Ang,
Low risk, low beta, manner, over-allocating to riskier stocks.
risky stocks, short positions credit, FX, Hodrick, Xing, Zhang
low volatility… This generates a market anomaly that,
on the riskiest stocks. commodities. (2009); Frazzini and
overall, is beneficial to the least risky
Pedersen (2014).
securities (less susceptible to market
corrections).
Koijen, Moskowitz,
Pedersen, Vrugt (2016);
Benefit from the yield differential (rates, Long positions in high yield Equities, rates,
Gorton, Hayashi,
Carry coupons, dividends, etc.) between similar securities, short on low credit, FX,
Rouwenhorst (2012);
assets. yielding ones. commodities.
Brooks, Moskowitz
(2017).
Benefit from the outperformance of Greenblatt (2006);
Long high quality
companies that show superior quality, in Asness, Frazzini,
Quality compagnies, short low Equities.
terms of profitability, dividend distribution, Pedersen (2013); Novy-
quality compagnies.
credit quality, governance ... Marx (2014).
The structural demand for protection implies Coval, Shumway
Short straddle, delta hedged Equities, rates,
a structural difference between the levels of (2001); Ang, Israelov,
Short volatility by a long position in the credit, FX,
implied (higher) volatility and realized Sullivan, Tummala
underlying market. commodities.
volatility. (2018)
Short volatility future (delta
hedged) when the curve is in
Profit from the teem structure of the
Volatility carry contango. Reverse position Volatility
volatility curve.
when the curve is in
backwardation.
Trading Risk Premia

Long or short position in the

Exploit short-term market overreaction, Equities, rates,
underlying index in order to
generally measured by the difference credit, FX, Poterba, Summers
Mean reversion replicate the market
between short (daily) realized volatility and commodities, (1988)
sensitivity (delta) induced by
longer-term volatility (one or two weeks). volatility.
a variance swap.
Long positioning on
Equities, rates,
securities with positive trend Moskowitz, Ooi, and
Trend following / Exploit trends in asset prices, similar CTA credit, FX,
and/or short ones with Pedersen (2012); Fung,
absolute momentum strategies. commodities,
negative trend. Directional Hsieh (2001)
volatility.
strategy (long or short bias).
Long and/or short positions
on the securities from the
Equities, rates,
Directional versions investment universe, Cf. the references of
Cf. academic risk premia. credit, FX,
of academic ARPs defined by their exposure academic risk premia.
commodities.
level to the underlying risk
factor.
Source: G. MONARCHA (2019), “An Introduction to Alternative Risk Premia”, Alternative Investment Analyst Review, Q1 2019.

Alternative Risk Premia | October 2019 Orion Financial Partners | p.11

 Disclaimer
This document is strictly confidential and for the exclusive attention of its recipients. If you receive this document by mistake,
thank you delete it and notify the sender immediately.

This document reflects the opinion of Orion Financial Partners SAS on the date of its publication, unless otherwise outside
contributors. This opinion is subject to change at any time without notice. This document is created for information only. Neither
the information nor analysis that is expressed in any way constitute an offer to sell or a solicitation and does not engage the
responsibility of Orion Financial Partners SAS. Orion Financial Partners SAS cannot be held liable for financial losses or any decision
made on the basis of the information contained in this document.

Orion Financial Partners SAS does not warrant the accuracy or completeness of the information sources, although these sources
are considered reliable. Orion Financial Partners SAS therefore does not engage its responsibility under the disclosure or use of
information contained in this publication.

Alternative Risk Premia | October 2019 Orion Financial Partners | p.12