VWR could be an extension of Sharpe Ratio, so before understanding VWR let’s know Sharpe Ratio.

**Let’s understand why and what is Sharpe ratio?**

Sharpe ratio helps you to know how much extra (or exceed) return you would have received if you had invested the same amount in risk-free assets i.e. FD, Treasury Bonds.

In very simple terms Sharpe ratio will tell you if that strategy is worth investing or taking that pain to execute it. Sharpe ratio in the range of 1-2 % is considered to be a good investment strategy.

The standard deviation helps to show how much the portfolio's return deviates from the expected return. The standard deviation also sheds light on the portfolio's volatility.

**Problem with Sharpe ratio?**

We took sample data from 2019 to 2020 the values returned for different time frames were

Time frame years: 2.8748

Time frame months: 0.7849

Time frame Week:0.6785

Time frame Days:0.6175

Here as we cut down the timeframe the Sharpe ratio also decreases, this is due to larger sample trying to fit in a small time frame are creating variance (difference between the maximum and minimum returns), high variance leads to a smaller Sharpe ratio

The problem that arises is if there is a trade that has a potential of higher return, every time a variance occurs the Sharpe ratio fells further down, and on another hand, if the trade is performing way less than the previous trade but has less or no variance will have a good Sharpe ratio which will give misleading results.

**Drawback of Sharpe Ratio:**

The average return takes into account both the positive, as well as the negative returns. The Periodic returns are not additive. Even it does not factor in the time interval of data used to calculate the Return (daily, weekly, or yearly etc.

It uses a risk-free rate, with a change in the risk-free rate, Sharpe value changes. It gives different values to different investors.

Low volatile stocks will have smaller Standard Deviation of returns (represent risk), in denominator moves Sharpe ratio hyperbolically. As variability (as σ) goes to zero, the ratio becomes infinite, regardless of its return.

It does not discriminate between the low-return, low-volatility stock, and the high-return, high-volatility stock.

**How can Variable Weighted Returns help here**?

Variable weighted returns or VWR is an extension of Sharpe ratio. It overcomes the ambiguity of Sharpe ratio by introducing timeframe independent returns these returns are always in annualized form and can be executed on different time frames it gives consistent values for consistent returns

**Deep-dive on VWR**

VWR uses normalized logarithmic return times a multiplier which ranges from 0 to 1, Here instead of having a rigid structure as of Sharpe ratio in VWR, investors can choose price variability tolerance as per their preference. It lets you choose a maximum acceptable standard deviation of prices, for example, some risk averse investors say a newbie investor might go with lower S.D and/or a lower risk tolerance (τ) of not more than 1.

Logarithmic returns remove timeframe issue and with the flexibility of selecting risk tolerance level gives investors gives required insights on strategy

Rnorm - normalized return (avg log return, normalized to simple annualized return)

σP - standard deviation of price differentials

σmax - maximum acceptable σP (investor limit)

τ - rate at which weighting falls with increasing variability (investor tolerance)

**Advantage of VWR**

VWR uses normalized logarithmic return which is representative of actual compounded returns rather than an Average of (arithmetic mean) of returns. It uses the mean logarithmic return to demonstrate the zero-variability baseline for a stock of the same total return over the period.

It calculates standard deviation on price variability rather than on returns.

It does not require comparison with a benchmark fund or risk-free rate.

It gives an annualized value that is consistent for each stock, across daily, weekly, or monthly granularity. Properly normalized returns would be identical for each sub-period length.

It takes into consideration the time frame for which return data is calculated.

why cant i find anything abt this on internet?