wbfdm 1.43.1__py2.py3-none-any.whl

wbfdm/models/instruments/instrument_prices.py

@@ -0,0 +1,544 @@
+from contextlib import suppress
+from decimal import Decimal
+
+import pandas as pd
+from celery import shared_task
+from django.db import models, transaction
+from django.db.models import (
+    Case,
+    DecimalField,
+    Exists,
+    ExpressionWrapper,
+    F,
+    OuterRef,
+    Q,
+    QuerySet,
+    Subquery,
+    Value,
+    When,
+)
+from django.db.models.signals import post_save
+from django.dispatch import receiver
+from wbcore.contrib.currency.models import CurrencyFXRates
+from wbcore.contrib.io.mixins import ImportMixin
+from wbcore.models import DynamicDecimalField, DynamicFloatField, DynamicModel, WBModel
+from wbcore.signals import pre_merge
+from wbfdm.analysis.financial_analysis.financial_statistics_analysis import (
+    FinancialStatistics,
+)
+from wbfdm.import_export.handlers.instrument_price import InstrumentPriceImportHandler
+
+from .mixin.financials_computed import InstrumentPriceComputedMixin
+
+
+class ValidPricesQueryset(QuerySet):
+    def filter_only_valid_prices(self) -> QuerySet:
+        """
+        Filter the queryset to remove duplicate in case calculated and non-calculated prices are present for the same date/product/type
+        """
+        return self.annotate(
+            real_price_exists=Exists(
+                self.filter(
+                    instrument=OuterRef("instrument"),
+                    date=OuterRef("date"),
+                    calculated=False,
+                )
+            )
+        ).filter(Q(calculated=False) | (Q(real_price_exists=False) & Q(calculated=True)))
+
+    def annotate_market_data(self):
+        base_qs = ValidPricesQueryset(self.model, using=self._db)
+
+        return self.annotate(
+            currency_fx_rate_to_usd_rate=F("currency_fx_rate_to_usd__value"),
+            calculated_outstanding_shares=Subquery(
+                base_qs.filter(instrument=OuterRef("instrument"), calculated=True, date=OuterRef("date")).values(
+                    "outstanding_shares"
+                )[:1]
+            ),
+            internal_outstanding_shares=ExpressionWrapper(
+                Case(  # Annotate the parent security if exists
+                    When(
+                        outstanding_shares__isnull=False,
+                        then=F("outstanding_shares"),
+                    ),
+                    default=F("calculated_outstanding_shares"),
+                ),
+                output_field=DecimalField(max_digits=4),
+            ),
+            internal_market_capitalization=Case(
+                When(
+                    market_capitalization__isnull=True,
+                    then=ExpressionWrapper(
+                        F("internal_outstanding_shares") * F("net_value"),
+                        output_field=DecimalField(max_digits=4),
+                    ),
+                ),
+                default=ExpressionWrapper(F("market_capitalization"), output_field=DecimalField(max_digits=4)),
+            ),
+            internal_market_capitalization_usd=F("internal_market_capitalization") / F("currency_fx_rate_to_usd_rate"),
+            calculated_volume=Subquery(
+                base_qs.filter(instrument=OuterRef("instrument"), calculated=True, date=OuterRef("date")).values(
+                    "volume"
+                )[:1]
+            ),
+            internal_volume=ExpressionWrapper(
+                Case(  # Annotate the parent security if exists
+                    When(
+                        volume__isnull=False,
+                        then=F("volume"),
+                    ),
+                    default=F("calculated_volume"),
+                ),
+                output_field=models.FloatField(),
+            ),
+        )
+
+    def annotate_base_data(self):
+        return self.annotate(
+            currency_fx_rate_to_usd_rate=F("currency_fx_rate_to_usd__value"),
+            market_capitalization_usd=ExpressionWrapper(
+                F("market_capitalization") / F("currency_fx_rate_to_usd_rate"), output_field=DecimalField(max_digits=4)
+            ),
+            net_value_usd=F("net_value") / F("currency_fx_rate_to_usd_rate"),
+            gross_value_usd=F("gross_value") / F("currency_fx_rate_to_usd_rate"),
+            volume_usd=ExpressionWrapper(
+                F("volume") * F("net_value") / F("currency_fx_rate_to_usd_rate"), output_field=DecimalField()
+            ),
+            volume_50d_usd=ExpressionWrapper(
+                F("volume_50d") * F("net_value") / F("currency_fx_rate_to_usd_rate"), output_field=DecimalField()
+            ),
+            volume_200d_usd=ExpressionWrapper(
+                F("volume_200d") * F("net_value") / F("currency_fx_rate_to_usd_rate"), output_field=DecimalField()
+            ),
+        )
+
+    def annotate_security_data(self):
+        return self.annotate(
+            security=Case(  # Annotate the parent security if exists
+                When(instrument__parent__isnull=False, then=F("instrument__parent")),
+                default=F("instrument"),
+            ),
+            security_instrument_type_key=Case(  # Annotate the parent security if exists
+                When(
+                    instrument__parent__isnull=False,
+                    then=F("instrument__parent__instrument_type__key"),
+                ),
+                default=F("instrument__instrument_type__key"),
+            ),
+        )
+
+    def annotate_all(self):
+        return self.annotate_market_data().annotate_security_data().annotate_base_data()
+
+
+class InstrumentPriceManager(models.Manager):
+    def __init__(self, with_annotation: bool = False, *args, **kwargs):
+        self.with_annotation = with_annotation
+        super().__init__(*args, **kwargs)
+
+    def get_queryset(self) -> ValidPricesQueryset:
+        qs = ValidPricesQueryset(self.model)
+        if self.with_annotation:
+            qs = qs.annotate_all()
+        return qs
+
+    def filtered_by_instruments(self, instrument_queryset, *other_instruments):
+        return self.filter(models.Q(instrument__in=instrument_queryset) | models.Q(instrument__in=other_instruments))
+
+    def filter_only_valid_prices(self) -> QuerySet:
+        return self.get_queryset().filter_only_valid_prices()
+
+    def annotate_market_data(self) -> QuerySet:
+        return self.get_queryset().annotate_market_data()
+
+    def annotate_base_data(self) -> QuerySet:
+        return self.get_queryset().annotate_base_data()
+
+    def annotate_security_data(self) -> QuerySet:
+        return self.get_queryset().annotate_security_data()
+
+    def annotate_all(self) -> QuerySet:
+        return self.get_queryset().annotate_all()
+
+
+class AnnotatedInstrumentPriceManager(InstrumentPriceManager):
+    def get_queryset(self):
+        return super().get_queryset().annotate_market_data().annotate_base_data().annotate_security_data()
+
+
+class InstrumentPrice(
+    ImportMixin,
+    InstrumentPriceComputedMixin,
+    DynamicModel,
+    WBModel,
+):
+    import_export_handler_class = InstrumentPriceImportHandler
+    # Base fields
+    instrument = models.ForeignKey(
+        to="wbfdm.Instrument",
+        related_name="prices",
+        on_delete=models.PROTECT,
+        limit_choices_to=models.Q(children__isnull=True),
+        verbose_name="Instrument",
+        blank=True,
+        null=True,
+    )
+    date = models.DateField(verbose_name="Date")
+    calculated = models.BooleanField(default=False, verbose_name="Is Calculated")
+
+    net_value = models.DecimalField(max_digits=16, decimal_places=6, verbose_name="Value (Net)")
+    gross_value = DynamicDecimalField(
+        max_digits=16,
+        decimal_places=6,
+        verbose_name="Value (Gross)",
+    )  # TODO: I think we need to remove this field that is not really used here.
+
+    outstanding_shares = DynamicDecimalField(
+        decimal_places=4,
+        max_digits=16,
+        verbose_name="Outstanding Shares",
+        help_text="The amount of outstanding share for this instrument",
+    )
+    outstanding_shares_consolidated = DynamicDecimalField(
+        decimal_places=4,
+        max_digits=16,
+        verbose_name="Outstanding Shares (Consolidated)",
+        help_text="The amount of outstanding share for this instrument",
+    )
+    ########################################################
+    #                   ASSET STATISTICS                   #
+    ########################################################
+
+    volume = models.FloatField(
+        null=True,
+        blank=True,
+        verbose_name="Volume",
+        help_text="The Volume of the Asset on the price date of the Asset.",
+    )
+
+    volume_50d = DynamicFloatField(
+        verbose_name="Average Volume (50 Days)",
+        help_text="The Average Volume of the Asset over the last 50 days from the price date of the Asset.",
+    )
+
+    volume_200d = models.FloatField(
+        null=True,
+        blank=True,
+        verbose_name="Average Volume (200 Days)",
+        help_text="The Average Volume of the Asset over the last 200 days from the price date of the Asset.",
+    )
+    market_capitalization = models.FloatField(
+        null=True,
+        blank=True,
+        verbose_name="Market Capitalization",
+        help_text="The Market Capitalization of the Asset the price date of the Asset.",
+    )
+    market_capitalization_consolidated = models.FloatField(
+        null=True,
+        blank=True,
+        verbose_name="Market Capitalization (Consolidated)",
+        help_text="the consolidated market value of a company in local currency.",
+    )
+
+    currency_fx_rate_to_usd = models.ForeignKey(
+        to="currency.CurrencyFXRates",
+        related_name="instrument_prices",
+        on_delete=models.PROTECT,
+        blank=True,
+        null=True,
+        verbose_name="Instrument Currency Rate",
+        help_text="Rate to between instrument currency and USD",
+    )
+
+    # Statistics
+    lock_statistics = models.BooleanField(
+        default=False,
+        help_text="If True, a save will not override the beta, correlation and sharpe ratio",
+    )
+    sharpe_ratio = models.FloatField(blank=True, null=True, verbose_name="Sharpe Ratio")
+    correlation = models.FloatField(blank=True, null=True, verbose_name="Correlation")
+    beta = models.FloatField(blank=True, null=True, verbose_name="Beta")
+    annualized_daily_volatility = models.FloatField(blank=True, null=True, verbose_name="Annualized Volatility")
+
+    created = models.DateTimeField(auto_now_add=True, verbose_name="Created")
+    modified = models.DateTimeField(
+        verbose_name="Modified",
+        auto_now=True,
+    )
+    # custom_beta_180d = DynamicFloatField(verbose_name="Custom Beta (180 days)")
+    # custom_beta_1y = DynamicFloatField(verbose_name="Custom Beta (1 Years)")
+    # custom_beta_2y = DynamicFloatField(verbose_name="Custom Beta (2 Years)")
+    # custom_beta_3y = DynamicFloatField(verbose_name="Custom Beta (3 Years)")
+    # custom_beta_5y = DynamicFloatField(verbose_name="Custom Beta (4 Years)")
+    #
+    # # Performances fields
+    # performance_1d = DynamicDecimalField(
+    #     verbose_name="Performance 1D", help_text="Performance 1D", max_digits=16, decimal_places=6
+    # )
+    # performance_7d = DynamicDecimalField(
+    #     verbose_name="Performance (1W)", help_text="Performance 7 days rolling", max_digits=16, decimal_places=6
+    # )
+    # performance_30d = DynamicDecimalField(
+    #     verbose_name="Performance (1M)", help_text="Performance 30 days rolling", max_digits=16, decimal_places=6
+    # )
+    # performance_90d = DynamicDecimalField(
+    #     verbose_name="Performance (1Q)", help_text="Performance 90 days rolling", max_digits=16, decimal_places=6
+    # )
+    # performance_365d = DynamicDecimalField(
+    #     verbose_name="Performance (1Y)", help_text="Performance 365 days rolling", max_digits=16, decimal_places=6
+    # )
+    # performance_ytd = DynamicDecimalField(
+    #     verbose_name="Performance (YTD)", help_text="Performance Year-to-date", max_digits=16, decimal_places=6
+    # )
+    # performance_inception = DynamicDecimalField(
+    #     verbose_name="Performance (Inception)",
+    #     help_text="Performance since inception",
+    #     max_digits=16,
+    #     decimal_places=6,
+    # )
+
+    objects = InstrumentPriceManager()
+    annotated_objects = InstrumentPriceManager(with_annotation=True)
+
+    class Meta:
+        verbose_name = "Instrument Price"
+        verbose_name_plural = "Instrument Prices"
+        constraints = [
+            models.CheckConstraint(
+                check=~models.Q(date__week_day__in=[1, 7]),
+                name="%(app_label)s_%(class)s_weekday_constraint",
+            ),
+            models.UniqueConstraint(fields=["instrument", "date", "calculated"], name="unique_price"),
+        ]
+        indexes = [
+            models.Index(
+                name="fdm_instrumentprice_base_idx",
+                fields=["calculated", "date", "instrument"],
+            ),
+            models.Index(
+                name="fdm_instrumentprice_idx1",
+                fields=["calculated", "instrument"],
+            ),
+            models.Index(
+                name="fdm_instrumentprice_idx2",
+                fields=["instrument"],
+            ),
+        ]
+
+    @property
+    def _net_value_usd(self):
+        if self.currency_fx_rate_to_usd:
+            return getattr(self, "net_value_usd", self.net_value * self.currency_fx_rate_to_usd.value)
+
+    def save(self, *args, **kwargs):
+        if not self.currency_fx_rate_to_usd:
+            with suppress(CurrencyFXRates.DoesNotExist):
+                self.currency_fx_rate_to_usd = CurrencyFXRates.objects.get(
+                    date=self.date, currency=self.instrument.currency
+                )
+
+        if self.market_capitalization_consolidated is None:
+            self.market_capitalization_consolidated = self.market_capitalization
+
+        super().save(*args, **kwargs)
+
+    def __str__(self):
+        return f"{self.instrument.name}: {self.net_value} {self.date:%d.%m.%Y}"
+
+    @property
+    def previous_price(self):
+        """Returns the previous instrument prices if one exists or None
+
+        Returns:
+            instrument.InstrumentPrice | None -- Previous InstrumentPrice
+        """
+        try:
+            return InstrumentPrice.objects.filter(
+                instrument=self.instrument, date__lt=self.date, calculated=self.calculated
+            ).latest("date")
+        except InstrumentPrice.DoesNotExist:
+            return None
+
+    @property
+    def next_price(self):
+        """Returns the next instrument prices if one exists or None
+
+        Returns:
+            instrument.InstrumentPrice | None -- Next InstrumentPrice
+        """
+        try:
+            return self.instrument.prices.filter(date__gt=self.date, calculated=self.calculated).earliest("date")
+        except InstrumentPrice.DoesNotExist:
+            return None
+
+    @property
+    def shares(self):
+        """Returns the number of shares of a instrument
+
+        The number of shares are the sum of all customer trades
+
+        Returns:
+            int -- Shares
+        """
+        return self.instrument.total_shares(self.date)
+
+    @property
+    def valid_outstanding_shares(self):
+        prices = self.instrument.prices.filter(date=self.date, outstanding_shares__isnull=False).order_by("calculated")
+        return prices.last().outstanding_shares if prices.exists() else self.shares
+
+    @property
+    def nominal_value(self):
+        """Returns the nominal value of a instrument
+
+        The nominal value is the number of current shares multiplied by the share price of a instrument
+
+        Returns:
+            int -- Nominal Value
+        """
+        return self.instrument.nominal_value(self.date)
+
+    def compute_and_update_statistics(self, min_period: int = 20):
+        prices_df = (
+            pd.DataFrame(
+                InstrumentPrice.objects.filter_only_valid_prices()
+                .filter(instrument=self.instrument, date__lte=self.date)
+                .values_list("date", "net_value"),
+                columns=["date", "net_value"],
+            )
+            .set_index("date")
+            .sort_index()["net_value"]
+        )
+        if not prices_df.empty and prices_df.shape[0] >= min_period:
+            financial_statistics = FinancialStatistics(prices_df)
+            min_date = prices_df.index.min()
+            if risk_free_rate := self.instrument.primary_risk_instrument:
+                risk_free_rate_df = (
+                    pd.DataFrame(
+                        risk_free_rate.valuations.filter(date__gte=min_date, date__lte=self.date).values_list(
+                            "date", "net_value"
+                        ),
+                        columns=["date", "net_value"],
+                    )
+                    .set_index("date")
+                    .sort_index()["net_value"]
+                )
+                if sharpe_ratio := financial_statistics.get_sharpe_ratio(risk_free_rate_df):
+                    self.sharpe_ratio = sharpe_ratio
+            if benchmark := self.instrument.primary_benchmark:
+                benchmark_df = (
+                    pd.DataFrame(
+                        benchmark.valuations.filter(date__gte=min_date, date__lte=self.date).values_list(
+                            "date", "net_value"
+                        ),
+                        columns=["date", "net_value"],
+                    )
+                    .set_index("date")
+                    .sort_index()["net_value"]
+                )
+                if (beta := financial_statistics.get_beta(benchmark_df)) is not None:
+                    self.beta = beta
+                if (correlation := financial_statistics.get_correlation(benchmark_df)) is not None:
+                    self.correlation = correlation
+
+            self.annualized_daily_volatility = financial_statistics.get_annualized_daily_volatility()
+
+    @classmethod
+    def subquery_closest_value(
+        cls,
+        field_name,
+        val_date=None,
+        date_name="date",
+        instrument_pk_name="instrument__pk",
+        date_lookup="lte",
+        order_by="-date",
+        calculated_filter_value=False,
+    ):
+        index_filter_params = {}
+        if calculated_filter_value is not None:
+            index_filter_params["calculated"] = calculated_filter_value
+        if not val_date and date_name:
+            index_filter_params[f"date__{date_lookup}"] = models.OuterRef(date_name)
+        elif val_date:
+            index_filter_params[f"date__{date_lookup}"] = val_date
+        index_filter_params["instrument"] = models.OuterRef(instrument_pk_name)
+        qs = cls.objects.filter(**index_filter_params).filter(**{f"{field_name}__isnull": False})
+        return models.Subquery(qs.order_by(order_by).values(field_name)[:1])
+
+    @classmethod
+    def annotate_sum_shares(cls, queryset, val_date, date_key="date"):
+        """
+        Efficient way to annotate sum of shares without destroying indexing
+        """
+        return queryset.annotate(
+            sum_shares_calculated=InstrumentPrice.subquery_closest_value(
+                "outstanding_shares",
+                date_name=date_key,
+                instrument_pk_name="pk",
+                date_lookup="lte",
+                order_by="-date",
+                calculated_filter_value=True,
+            ),  # We get the last price whose outstanding_shares is not none and date below date_key
+            sum_shares_real=InstrumentPrice.subquery_closest_value(
+                "outstanding_shares",
+                val_date=val_date,
+                date_name=date_key,
+                instrument_pk_name="pk",
+                date_lookup="exact",
+                calculated_filter_value=False,
+            ),
+            sum_shares=Case(
+                When(sum_shares_real__isnull=False, then=F("sum_shares_real")),
+                When(sum_shares_calculated__isnull=False, then=F("sum_shares_calculated")),
+                default=Value(Decimal(0)),
+                output_field=DecimalField(),
+            ),
+        )
+
+    @classmethod
+    def get_endpoint_basename(cls):
+        return "wbfdm:price"
+
+    @classmethod
+    def get_representation_value_key(cls):
+        return "id"
+
+    @classmethod
+    def get_representation_label_key(cls):
+        return "{{instrument} {{date}}"
+
+    @classmethod
+    def get_representation_endpoint(cls):
+        return "wbfdm:price-list"
+
+
+@receiver(post_save, sender="currency.CurrencyFXRates")
+def rate_creation(sender, instance, created, raw, **kwargs):
+    if not raw and created:
+        transaction.on_commit(lambda: update_currency_fx_rate_from_created_rate.delay(instance.id))
+
+
+@receiver(pre_merge, sender="wbfdm.Instrument")
+def pre_merge_instrument(sender: models.Model, merged_object, main_object, **kwargs):
+    """
+    Simply reassign the prices of the merged instrument to the main instrument if they don't already exist for that day, otherwise simply delete them
+    """
+    merged_object.prices.annotate(
+        already_exists=Exists(
+            InstrumentPrice.objects.filter(
+                calculated=OuterRef("calculated"), instrument=main_object, date=OuterRef("date")
+            )
+        )
+    ).filter(already_exists=True).delete()
+    merged_object.prices.update(instrument=main_object)
+
+
+@shared_task(queue="portfolio")
+def update_currency_fx_rate_from_created_rate(rate_id):
+    currency_rate = CurrencyFXRates.objects.get(id=rate_id)
+    for price in InstrumentPrice.objects.filter(
+        instrument__currency=currency_rate.currency, date=currency_rate.date, currency_fx_rate_to_usd__isnull=True
+    ):
+        price.currency_fx_rate_to_usd = currency_rate
+        price.save()