A Handy-dandy Reference to the “Master Spreadsheet” (“GEX+ CSV” on the GammaVol page)

The spreadsheet updates around 5:30am every morning after a trading day.

Each column is described below.

DATE

Date

Date is presented in format YYYY-MM-DD.

SPX

S&P 500 closing price

Value of the S&P 500 index at close.

N.B., this is not the S&P 500 “Total Return” index.

CHG(%)

1-day S&P 500 change

The close-to-close S&P 500 1-day change, in percent.

more

GEX

Gamma exposure
The gamma exposure of all existing S&P 500
(SPX) options, measured in thousands of dollars
per SPX point. This particular GEX implementation
uses Dealer Directional Open Interest (see DDOI in GEXto CHG
Appendix) to determine which options are held
short or long by dealers. Positive GEX means
dealers are providing liquidity to (stabilizing) the
market when spot price moves; negative GEX
means dealers are taking liquidity from
(destabilizing) the market when spot price moves.

Example:

The GEX value on 2004-01-02 was 64521.235.

This means that, for every 1-point down (up) in the
S&P 500, option dealers would need to buy (sell)
$64,521,235 worth of S&P 500 index exposure. So
if the index fell from 1108.48 to 1107.48, dealers
would be buying $64,521,235 of the S&P 500 to
re-hedge their deltas.
destabilizing stabilizing

VEX

Vanna exposure

The vanna exposure of all existing S&P 500 (SPX) options, measured in thousands of dollars per SPX
point, where each 0.10% move in SPX is assumed, for simplicity, to result in an anti-correlated 1.00%
move (10x) in option implied volatility. This seeks to measure the impact of dealers' delta re-hedging
activity with respect to changes in implied volatility (vanna) rather than with respect to changes in spot
price (gamma). As with GEX above, this VEX implementation uses Dealer Directional Open Interest (see
DDOI in Appendix). When VEX is positive,
dealers are providing liquidity to (stabilizing) the
market when implied volatility rises; when VEX
VEXto CHG is negative, dealers are taking liquidity from
(destabilizing) the market when implied
volatility rises.

Example:

The VEX value on 2004-01-02 was 72414.74.

This means that, for every 1-point down (up) in
the S&P 500, option dealers would need to buy
(sell) $72,414,740 worth of S&P 500 index
exposure, assuming that a 1-point move down
(up) in SPX results in -10x the percentage
change in vols. So if the index fell from 1108.48
to 1107.48 (-0.000902136%), we would
assume that implied volatility would rise
0.0902136% (-0.000902136% * -10) across the
board, and that dealers would as a result be
buying $72,414,740 of the S&P 500 to re-
destabilizing stabilizing hedge their deltas.

GEX+

Gamma exposure + vanna exposure

The straight sum of GEX and VEX. Since

GEX to CHG
GEX and VEX are the two dealer delta
sensitivities that have a large, persistent,
and immediate impact on current S&P 500
liquidity, we want to be able to view their GIV 100
combined market impact. Because we
made the effort to denominate VEX in GIV 20
"dollars per SPX point" -- the same units GIV 5
as GEX -- we are able to simply sum GEX
and VEX to achieve this result (GEX+).

GIV

Gamma-implied volatility

The 1-day historical volatility associated

with the current GEX+ value. When GEX+
is high, GIV will be low; when GEX+ is low,
GIV will be high.

GIV(MAD%)

Gamma-implied volatility (MAD)

The mean absolute deviation (see MAD in Appendix) equivalent of GIV. I.e., the 1-day average S&P 500
move associated with the current level of GEX+.

GN MAD GIV X
IFF IF

NPD

Net put delta

The combined net customer delta of all of the

day's SPX put option trades. E.g., a -20.00 NPD to Volatility
NPD means that the combined daily put
activity netted out to customers buying 20
deltas of puts from dealers; and a 20.00 NPD

I
means that it netted out to customers selling
20 deltas of puts to dealers. When NPD is in
the "shallow negatives" (-5.00, 0.00), that I
means that customers are mostly swapping stder
s
deltas with each other, and more new option o
positions are held without dealers as an mean
median
intermediary. Since option customers, in I
aggregate, hedge less frequently than
dealers, this increases aggregate market risk
and fragility, and large moves in market
volatility are likelier to occur (high vol of vol).
NPD

Example:

When NPD is -10, that means people are buying lots of puts, which tends to stabilize the market:
Investors are more "insured" (sentiment bullish) option customers will "monetize" their puts if SPX falls
(technical bullish), and if SPX doesn't fall, the put positions will decay and push the index up (flows
bullish). When NPD is 0, neither dealers nor customers will be stabilizing the market, and volatility (VIX)
can expand dramatically.

VGR

Vanna-gamma ratio

The ratio of SPX option customers' vanna exposure to their gamma exposure, where "customer"
exposure is any option exposure that is not held by dealers. E.g., for every dealer position (every
contract in DDOI), there is one customer exposure, and for every position held between two customers
(every contract in OI that is not in DDOI), there are two customer exposures. When vanna becomes
larger relative to gamma, (-5.00, 0.00), customers have more exposure to changes in market volatility,
and volatility is more likely to increase.

GR to Volatility Example:

When VGR is -10, that means customers'

exposure to gamma is 1000% higher than
vanna (stable). When VGR is -2, customers'
exposure to gamma is just 200% higher than
stder vanna (unstable). In the former case, VIX will
tend to fall, and in the latter case, VIX will
tend to rise. When VGR is positive, that
P median
invariably means customers' exposure to
I mean
gamma is low, and market volatility tends to
already be high. In these situations, NPD (a
measure of current put flows) becomes more
important.

VGR

VIX

Cboe Volatility Index

Closing value of VIX.

VIX(MAD%)

Cboe Volatility Index (1-day MAD)

The 1-day mean absolute deviation (see MAD in Appendix) equivalent of VIX. I.e., the 1-day average
S&P 500 move associated with the current level of VIX.

CR(x)

Crash-risk multiple

The gamma-implied volatility at 10% below SPX over the 30-day market-implied volatility at 10%
below SPX. E.g., if the -10% GIV is 40 and the -10% market IV is 20, then the crash-risk multiple is
2.00x. The higher the multiple, the greater the risk that a market correction (-10%) can spiral out of
control and become a liquidity crisis, facilitating a crash. Historically, a CR(x) of 3.00x or higher may be
a cause for concern.

0
agree
SPX v at
andvolatility
VIX gwthatlogowhichbe
would means
is1.00
28 G

CRG

Example:

If customers have been selling put options to dealers, and these puts have been accumulating well
below the market, there is a chance that, if SPX falls enough, dealers will end up having negative vanna
exposure (-VEX), which can precipitate a crash.

See "The Implied Order Book" for more context on how this causes crashes.

SU, MO, MID

Support, Momentum, Midpoint

The "SuMo bands." I.e., the -0.50 (Support), +0.25 (Midpoint), and +1.00 (Momentum) standard
deviation bands associated with the implied volatility of the prior SPX close. The Support and
Momentum bands are associated with bullish intraday action, and the Midpoint is associated with
bearish intraday action.

See the 2/28/21 Sunday note for more on how it works.

VIBE

Volatility-Implied Bullish Expectation

A measure of the relative changes in volatility between the S&P 500 (SPY) and the front-month VIX
futures (UVXY). Data range is [0, 100]. If SPY volatility is rising relative to UVXY volatility, that means
VIBE will be >50. If SPY volatility is falling relative to UVXY volatility, VIBE will be <50. Low VIBE is
bullish over the next month; high VIBE is bearish.

Appendix

MAD

Mean absolute deviation

The average (mean) move implied by an option's price/volatility. I.e., the magnitude for which 50% of
occurrences will end less than, and 50% will end greater than. The true mathematical basis for an
option's price. For most applications, MAD is preferable to standard deviation (STD)

ITF IF

MAD VOL x

OI
STD 0.7978
Open interest

The number of option contracts that exist, broken down by expiration, strike, and type. Data from the
Option Clearing Company (OCC).

DDOI

Dealer directional open interest

The number of option contracts that are held by option dealers, and the direction in which those
contracts are held. When dealers are short the option, the DDOI is negative; when dealers are long the
option, the DDOI is positive. DDOI is created by assessing trade direction of all option volume
throughout the day, then comparing that volume to subsequent change in open interest.

Example:

The 1-year 3000-strike put (OTM) was just listed. OI and DDOI are zero. On the first day of trading, 100
contracts appear to have been bought, and the OCC says that 100 contracts now exist in OI. At the
end of the first day, OI is 100 and DDOI is -100 (dealer is short the contracts). On the second day, 1000
contracts were sold, and reported OI changed to 900. We guess that those 100 contracts were sold to
the same dealer, and that 100 of those contracts offset the -100 in existing DDOI, leaving the dealer net
long 900 contracts. DDOI is now +900. On day three, we see two 500-contract trades occur -- one a
sale and one a purchase. OI goes up to 1400 (+500). We don't know if there's a new dealer involved, or
if customers traded with each other, but even though OI went up, we don't think dealers' exposures
went up any, so DDOI remains at +900.